Market Assessment of Public Sector Information
MARKET ASSESSMENT OF
PUBLIC SECTOR INFORMATION
Written by Deloitte
MAY 2013
Market Assessment of Public Sector Information
A report by Deloitte for the Department for Business, Innovation and Skills
Deloitte
Stonecutter Court
1 Stonecutter Street
London
EC4A 4TR
Web: UUwww.deloitte.co.uk
The views expressed in this report are the authors’ and do not necessarily reflect those of the
Department for Business, Innovation and Skills.
Department for Business, Innovation and Skills
1 Victoria Street
London SW1H 0ET
www.gov.uk/bis
URN BIS/13/743
May 2013
2
Market Assessment of Public Sector Information
Contents
Acknowledgements ........................................................................................................................ 9 Executive Summary...................................................................................................................... 10 Key findings................................................................................................................................. 10 Policy context .............................................................................................................................. 12 Defining public sector information and the market ...................................................................... 13 Public sector information dimensions...................................................................................... 13 Public sector information holders and customers ................................................................... 14 The value of public sector information today............................................................................... 16 The link between public sector information and growth .......................................................... 16 Generating value from public sector information..................................................................... 16 Value from different public sector information datasets .......................................................... 17 Understanding how public sector information creates value in the UK ................................... 19 Estimates of the value of public sector information and value to PSIHs ................................. 21 Barriers to realising the full potential of public sector information in the UK ............................... 24 A taxonomy of barriers ............................................................................................................ 24 Legislative barriers .................................................................................................................. 25 Economic barriers ................................................................................................................... 27 Access barriers ....................................................................................................................... 32 Summary of key research questions answered .......................................................................... 35 1. Introduction and approach................................................................................................... 42 Project scope and objectives ...................................................................................................... 42 Approach..................................................................................................................................... 43 Data sources ............................................................................................................................... 43 3
Market Assessment of Public Sector Information
Structure of this report................................................................................................................. 44 2. Public sector information definitions and market definition ............................................ 46 Definition of public sector information ......................................................................................... 46 Definition of the public sector ...................................................................................................... 47 The nature of public sector information....................................................................................... 48 The availability of public sector information ............................................................................ 49 The complexity of public sector information ............................................................................ 50 The speed at which public sector information is updated ....................................................... 50 Public sector information as a by-product or purpose-collected ............................................. 50 The content of public sector information ................................................................................. 51 The format of public sector information................................................................................... 52 The distribution channels of public sector information ............................................................ 54 The cost of public sector information ...................................................................................... 55 Market definition.......................................................................................................................... 55 Chapter summary........................................................................................................................ 57 3. The supply and demand sides of the public sector information market ......................... 59 Supply side of the market............................................................................................................ 59 Supply of datasets nationally .................................................................................................. 59 Supply of datasets locally........................................................................................................ 64 Trading Funds ......................................................................................................................... 66 Attitudes and policies of public sector information holders ..................................................... 69 Demand side of the market ......................................................................................................... 70 Public sector information use and re-use................................................................................ 71 Consumer perceptions of public sector information and unmet demand ................................ 75 Public sector information customer types ............................................................................... 80 4
Market Assessment of Public Sector Information
Different business models of public sector information intermediaries ................................... 86 Data scientists......................................................................................................................... 91 Chapter summary........................................................................................................................ 93 4. Public sector information regulatory landscape................................................................ 95 Policy and regulatory stakeholders ............................................................................................. 95 The legal framework for publishing and sharing public sector information ................................. 96 Public sector information licensing and the regulatory framework .......................................... 97 The Data Protection Act ........................................................................................................ 101 The European dimension .......................................................................................................... 103 The European Directive on the Re-Use of Public Sector Information ................................... 103 The proposed European Data Protection Regulation ........................................................... 103 The INSPIRE Directive.......................................................................................................... 104 Chapter summary...................................................................................................................... 105 5. The current value of the public sector information market ............................................ 106 Understanding the value of public sector information ............................................................... 107 Modelling approach and assumptions....................................................................................... 108 Stage 1: consumer and producer surplus approach – the welfare approach........................ 108 Stage 2: indirect and induced value of public sector information .......................................... 109 Narrow economic value estimates ............................................................................................ 110 Wider societal value estimates.................................................................................................. 110 ‘Blue sky’ opportunities ............................................................................................................. 116 Detection and Identification of Infectious Diseases (2006) ................................................... 116 Tackling Obesity (2007) ........................................................................................................ 116 Global Food and Farming (2011) .......................................................................................... 117 Computer Based Financial Trading (2012) ........................................................................... 117 5
Market Assessment of Public Sector Information
Reducing the Risks of Future Disasters (2012) .................................................................... 117 The Future of Identity (2013)................................................................................................. 117 Chapter summary...................................................................................................................... 118 6. Barriers and policy implications........................................................................................ 120 Generating value....................................................................................................................... 120 A taxonomy of barriers .............................................................................................................. 121 Legislative barriers ................................................................................................................ 122 Economic barriers ................................................................................................................. 128 Access barriers ..................................................................................................................... 133 Chapter summary...................................................................................................................... 140 7. Conclusion........................................................................................................................... 141 Closing thoughts ....................................................................................................................... 141 Appendix 1: Glossary ................................................................................................................. 143 Appendix 2: Acronyms used in this report .............................................................................. 154 Appendix 3: Literature Review .................................................................................................. 156 Overview ................................................................................................................................... 156 The economic impact ................................................................................................................ 157 The economic contribution of public sector information ........................................................ 158 The costs of providing public sector information ................................................................... 161 Providing value for money from public sector information .................................................... 161 Summary............................................................................................................................... 162 The broader impact ................................................................................................................... 163 Maximising the impact of public sector information................................................................... 164 Pricing models....................................................................................................................... 164 The impact of zero or marginal cost pricing on innovation.................................................... 166 6
Market Assessment of Public Sector Information
Case studies ............................................................................................................................. 168 Public sector efficiencies....................................................................................................... 168 Business innovation .............................................................................................................. 169 Country case studies............................................................................................................. 170 Appendix 4: Further statistics ................................................................................................... 174 Companies House..................................................................................................................... 174 Land Registry ............................................................................................................................ 174 The Met Office........................................................................................................................... 175 Ordnance Survey ...................................................................................................................... 179 Appendix 5: Empirical methodology......................................................................................... 181 General approach and limitations ............................................................................................. 181 Definition of value...................................................................................................................... 182 Methodologies considered ........................................................................................................ 183 The quantification approach for the current value of public sector information......................... 186 Stage 1: consumer and producer surplus approach – the welfare approach........................ 187 Stage 2: indirect and induced value of public sector information .......................................... 193 Appendix 6: Transport sector case study ................................................................................ 194 Overview ................................................................................................................................... 194 Information released by Transport for London .......................................................................... 194 Accessing information ........................................................................................................... 195 Usage levels.......................................................................................................................... 196 Benefits ................................................................................................................................. 197 Quantification of benefits........................................................................................................... 198 TfL ‘lost customer hours’ data ............................................................................................... 198 Apps using TfL data .............................................................................................................. 198 7
Market Assessment of Public Sector Information
The model ............................................................................................................................. 200 Findings................................................................................................................................. 200 Summary............................................................................................................................... 202 Information relating to transport beyond London ...................................................................... 204 Appendix 7: Further case studies ............................................................................................. 215 The healthcare and life science sector.................................................................................. 215 The public sector................................................................................................................... 221 Opportunities in the local public sector ................................................................................. 222 Appendix 8: Government Officials informal consultation....................................................... 226 Consultation questions.............................................................................................................. 226 Background ........................................................................................................................... 226 Consultation responses............................................................................................................. 228 Appendix 9: Bibliography .......................................................................................................... 230 Academic Papers & Reports ..................................................................................................... 230 Departmental Open Data Strategies ......................................................................................... 233 Other Documents ...................................................................................................................... 234 8
Market Assessment of Public Sector Information
Acknowledgements
This report has been produced by Deloitte between October 2012 and March 2013. The lead
authors are Andrew Tong, Haris Irshad and Daniel Revell Ward.
The Deloitte team acknowledges the support of Stephan Shakespeare and the Data Strategy
Board, the Cabinet Office, the Open Data Institute, other public sector bodies and the wider open
data community who have contributed their time, data and comments to this report. In particular,
the team would like to express its thanks to the different start-up companies and established
companies who have contributed to the case study analyses.
9
Market Assessment of Public Sector Information
Executive Summary
Key findings
This is the first UK-wide market assessment of public sector information 1 . It spans the use and reuse of public sector information at the UK-level, regionally and locally by a wide range of
businesses, civil society groups, government and members of the general public. The aim of this
market assessment is to establish a robust evidence base on its value and to highlight the policy
implications flowing from an examination of how public sector information could be utilised further.
The research has covered three broad thematic areas:

definitions of public sector information and its characteristics;

how public sector information is used and re-used inside and outside of government; and

barriers to fully exploiting the value of public sector information, including issues around
competitiveness, funding and regulation.
The research has used a combination of literature reviews, analysis of secondary sources,
stakeholder interviews and case studies. The key research findings are summarised below.
Key research findings
1

This report has estimated that the value of public sector information to consumers,
businesses and the public sector in 2011/12 was approximately £1.8 billion (2011 prices).
o This is a mid-point estimate, with the sensitivity analyses giving a range between £1.2
billion and £2.2 billion

However, the use and re-use of public sector information has much larger downstream
impacts affecting all areas of society beyond the direct customer.

The UK is a global leader in releasing public sector information for use by its citizens,
businesses and policymakers. Through initiatives such as the data.gov.uk, the Open Data
User Group, Departmental Data Strategies, as well as the establishment of the Data Strategy
Board and Open Data Institute, the UK has taken significant steps to creating a worldclass public data infrastructure.
o While there is no central figure on the number of public sector information datasets
currently being made available, a review of selected data portals suggests the number
could exceed 37,500 from over 750 different publishers with over 2.5 million downloads
by October 2012

There is a link between the provision and use/re-use of public sector information and
economic growth. Public sector information is used by businesses, individuals and the
public sector to:
o stimulate innovation and develop new products and services;
o hold public service providers to account, promote democratic engagement and
foster greater transparency and better policymaking;
Public Sector Information covers the wide range of information that public sector bodies collect, produce, reproduce
and disseminate in many areas of activity while accomplishing their public tasks.
The terms data, information and knowledge are frequently used for overlapping concepts. The main difference is in the
level of abstraction being considered. Data is a broad term, embracing others, but is often the lowest level of abstraction,
information is the next level and, finally, knowledge is the highest level. Unless otherwise stated, this report uses the term
‘data’ to include ‘information’ and vice versa. Accordingly, a ‘dataset’ in this context will be a collection of ‘data’ and/or
‘information’.
Market Assessment of Public Sector Information
o
o
reduce barriers to entry into markets and address information asymmetries; and
generate network effects that drive disruptive change by connecting increasing
numbers of consumers and businesses

The most popular, and potentially most valuable, datasets include geo-spatial,
environmental, transport, health and economic data, with the construction, real estate,
finance and insurance, public sector and arts, entertainment and recreation sectors being
some of the largest users and re-users of public sector information and open data.

Case studies examining the downstream impacts created by different organisations using
and re-using public sector information have shown the monetary value of some of these
benefits can be in the order of millions, if not billions. For example:
o publishing data on adult cardiac surgery is estimated to have reduced mortality rates,
which in turn has an economic value in excess of £400 million p.a.
o using live data from Transport for London in apps can save users time to the
economic value of between £15 million and £58 million p.a.

While an aggregate figure on the social value of public sector information is difficult to reach
without more information on the way data is used and how it then permeates society, on the
basis of conservative assumptions, it is estimated this figure could be in excess of £5
billion for 2011/12 (2011 prices).
o This estimate is likely to increase as public sector information is used more widely and in
more impactful ways.

Adding this social value estimate to the calculated value of public sector information to
consumers, businesses and the public sector, gives an aggregate estimate of between
£6.2 billion and £7.2 billion in 2011/12 (2011 prices). Future uses of public sector
information that have the potential to generate much more value include greater combining
of public and private sector information, exploiting the benefits of linked data,
embedding geospatial and location data across more and more products and services,
and more informed policymaking based on better utilised public sector information.

Using a taxonomy originally developed by the Data Strategy Board, this report has
considered a number of barriers that may be preventing the UK from fully realising the
benefits of public sector information:
o Legislative barriers: while current legislation, guidance and regulations are not
hindering the public sector information market themselves, there is often a perception
(right or wrong) that some legislation, guidance and regulations prevent datasets from
being released and shared thereby reducing the availability of datasets. The Open
Government Licence is widely seen as an effective means of improving the
availability of public sector information;
o Economic barriers: the issue of charging for public sector information datasets is
complex, with a lack of data from both providers and users/re-users making it hard to
conduct cost-benefit analyses on data release. While significant progress has been
made in making more datasets available, there remains a perception (actual and
perceived) that charging creates barriers to using and re-using public sector
information. In the short-term there are a number of opportunities for improving the
availability of certain datasets; and
o Access barriers: improvements continue to be made to reduce fragmentation and
improve the consistency of public sector information; however there remains scope
for improvements in the accessibility of public sector information datasets to the
general public. Equally, in some parts of the private and public sector, there is a skills
shortage preventing the effective extraction of value from public sector information.
11
Market Assessment of Public Sector Information
Policy context
As highlighted by Rt. Hon. Francis Maude MP in his foreword to the Open Data White Paper (June
2012), data is the twenty-first century’s raw material: “its value is in holding governments to
account; in driving choice and improvements in public services; and in inspiring innovation and
enterprise that spurs social and economic growth” 2 . Beginning with transparency statements
originally set in the Coalition Agreement 3 , the UK has continued to push forward with the release of
more public sector information, much of which meets the criteria of being ‘open data’. 4 Recent
developments include (these are not exhaustive):

the establishment of the Data Strategy Board (DSB 5 ) and the Open Data User Group 6
(ODUG) to create the maximum value for businesses and people across the UK from data held
by the Public Data Group 7 (PDG) members and beyond;

the establishment of the Transparency Board 8 to drive forward the Government’s
transparency agenda, set open data standards across the whole public sector and facilitate the
release of more public sector information;

the creation of the Open Data Institute (ODI), an independent, non-profit company funded by
the Government and the private sector to incubate, nurture and mentor new ideas using open
data (and the first organisation of its kind in the world);

the publication of new Open Standards Principles 9 for all government bodies to comply with
making IT more open, cheaper and better connected;

discussions at a European level on the revision of the Public Sector Information Directive
and reforms to data protection rules;

the publication of departmental open data commitments in their business plans;

the on-going implementation of the Government’s Transparency and Open Data agenda led
by Cabinet Office. This includes the publication of updates on the Government Digital
Strategy 10 providing details on how services are being moved online to achieve efficiencies
and how Cabinet Office is supporting improved digital capabilities across departments;

the completion of a competition awarding investment to Glasgow to demonstrate how a city of
the future could work by hosting the Technology Strategy Board’s Future City
Demonstrator 11 ; and

specific policy announcements such as:
2
Available at: www.gov.uk/government/publications/open-data-white-paper-unleashing-the-potential
Available at:
www.direct.gov.uk/prod_consum_dg/groups/dg_digitalassets/@dg/@en/documents/digitalasset/dg_187876.pdf
4
Open government data is a subset of public sector information. It is data which can be used, re-used and re-distributed
freely by anyone - subject only at most to the requirement to attribute and share-alike.
5
See www.gov.uk/data-strategy-board and the Terms of Reference at
www.gov.uk/government/uploads/system/uploads/attachment_data/file/32384/12-673-terms-reference-data-strategyboard-and-public-data-group.pdf
6
See http://data.gov.uk/odug
7
The members of the PDG are Ordnance Survey, the Met Office, Land Registry and Companies House.
8
See www.gov.uk/government/policy-advisory-groups/134
9
See www.gov.uk/government/news/government-bodies-must-comply-with-open-standards-principles
10
See www.gov.uk/government/news/update-on-government-digital-strategy-aligning-it-with-digital
11
See www.theodi.org/news/glasgow-wins-%C2%A324m-boost-after-recognising-%E2%80%9Crevolutionaryimpact%E2%80%9D-open-data
3
12
Market Assessment of Public Sector Information
o
the Open Data Immersion Programme 12 to accelerate innovative open data ideas and
develop market-ready businesses;
o
the expansion of the Technology Strategy Board’s Innovation Voucher programme
available to SMEs to use open data to help commercialise their ideas and develop new
products and prototypes; and
o
other funding for a DSB Breakthrough Fund and upgrades to Ordnance Survey
OpenData products 13 .
Together these developments are helping place the UK at the forefront of public sector information
and open data initiatives globally. Research by Deloitte suggests that the UK has the most page
views of any open data portal in the world 14 .
However, while the UK holds a number of strengths in the availability, analysis and value extraction
of public sector information, there are challenges in fully exploiting the potential value from public
sector information. There are logistical challenges around what data should be made available,
when, in which format and the choice of funding model to use. There also exist challenges around
ensuring businesses, civil society and the public sector itself are able to effectively use and re-use
public sector information so as to fully exploit its value.
The Government has recognised that these challenges may be becoming barriers and on 22
October 2012 launched the Shakespeare Review led by Stephan Shakespeare, Chair of the Data
Strategy Board. This review covers the entire breadth of the public sector information market
(current and future) in order to make recommendations to Ministers on how to widen access to
public sector information and consider new and innovative opportunities for open data. 15
As part of the Shakespeare Review, Deloitte has been commissioned by the Department of
Business, Innovation and Skills (BIS) on behalf of Stephan Shakespeare, to undertake an
independent assessment of the market for public sector information, in order to establish a robust
evidence base on its value and to highlight the policy implications flowing from an examination of
how public sector information could be utilised further.
Defining public sector information and the market
Public sector information dimensions
Broadly speaking the term ‘public sector information’ refers to data 16 and information 17 that the
public sector collects, produces, reproduces, publishes and disseminates in many areas of activity
while accomplishing their public task or other duties 18 . In some limited cases for particular public
sector information datasets 19 , there will also be private sector suppliers of data and information
that can be considered substitutes for public sector information.
12
See www.theodi.org/news/odi-launches-%C2%A3850k-scheme-create-businesses-open-data
See www.gov.uk/data-strategy-board#recent-news
14
Although it should be noted that page views do not always translate to use and re-use of data.
15
The full Terms of Reference can be found at: www.bis.gov.uk/assets/biscore/innovation/docs/i/12-1233-independentpublic-sector-information-review-draft-terms.pdf
16
Defined as qualitative or quantitative statements or numbers that are assumed to be factual and not the product of
analysis or interpretation.
17
Defined as outputs of processes that summarise, interpret or otherwise represent data to convey meaning. The term
data typically includes information and this document follows this convention.
18
The collection, production, reproduction, publication and dissemination of public sector information by the public sector
can be done through official channels or through third parties in the private and third sectors.
19
Defined as collections of data and information.
13
13
Market Assessment of Public Sector Information
Figure 1: Public sector information dimensions
Public sector information datasets differ across a number of dimensions:

whether the public sector information is made available to the general public and, if so, under what
conditions (if any);

the complexity of the dataset in terms of the number of records and variables, whether it is anonymised or
aggregated or is quantitative or qualitative;

how often the dataset is updated or replaced;

whether public sector information is generated or collected as part of an public body’s public task or
whether its generation is the result of other activities not related to public sector information;

the content of the public sector information dataset;

the electronic or non-electronic format of the dataset;

the ways in which the public sector information dataset is distributed; and

the cost of generating/collecting/maintaining/updating the public sector information dataset.
Source: Deloitte analysis
Public sector information holders and customers
Public sector bodies that generate, collect and disseminate public sector information (hereafter
referred to as Public Sector Information Holders or ‘PSIHs’) will themselves differ in size, role and
purpose and the extent to which, and how, they make the information and data available. Figure 2
below summarises.
Figure 2: Public sector information holders and routes to market
Source: Deloitte analysis. Note this illustrative is not exhaustive. For example, under the ‘20 year rule’, historical material
20
previously unavailable is made available to the public .
As Figure 2 illustrates, public sector information is generated by a plethora of different PSIHs. It
can be generated specifically (e.g. the Met Office collecting weather data), as a by-product of
20
In 2013, the government began its move towards releasing records when they were 20 years old instead of 30. See
www.nationalarchives.gov.uk/about/20-year-rule.htm#text for more details.
14
Market Assessment of Public Sector Information
other activities (e.g. salary data - so called exhaust data that is generated through the
performance of regular activities that are not data collection specific) or combined with third
party data (e.g. financial data analysed and combined with other statistics by the Bank of England
that is then used to inform economic forecasts).
While data on the total number of public sector information datasets is not recorded, a review of
selected data portals reveals there to be over 37,500 datasets currently being made available as
open data by PSIHs (this is likely to be an underestimate given a wealth of public sector
information is also available through other routes). Focusing on data.gov.uk, analysis shows that
the Office for National Statistics, Department for Communities and Local Government and the
Health and Social Care Information Centre – with the analysis showing that of a total of 781
different publishers (as of 27th February 2013), the top ten suppliers of public sector information
supplied over half of all datasets published on the website.
On the demand side of the market, customers come from the private sector, civil society,
individuals and government itself. Some customers will use the data as direct inputs into products,
services and research, while others will be repackaging the data for others to use – so-called
infomediaries.
The market for public sector information can therefore be summarised as shown in Figure 3 below.
Figure 3: Public sector information market
Source: Deloitte analysis
As Figure 3 shows, on the supply side, PSIHs consist of the public sector and private business –
though in practice the overwhelming majority will be supplied by the public sector. Intermediaries
take this information and data and host and repackage it for a wider audience – in some cases this
means augmenting the public sector information with other data elements. On the demand side,
final consumers (which can include PSIHs themselves) can use and re-use the public sector
information to develop new products, inform decision-making, improve research and make
efficiency savings. As public sector information is increasingly used and re-used, actors across the
supply, intermediary and demand sides will improve their data analysis skills, which can raise
competitiveness and drive economic growth.
15
Market Assessment of Public Sector Information
The value of public sector information today
The link between public sector information and growth
The economic importance of public sector information has increased radically with the spread of
new communication technologies, most notably the Internet, and the development of a ‘knowledge
economy’ in which value is generated through innovation in information and services. There is now
a considerable body of academic and other literature supporting the view that the greater
availability and accessibility 21 of public sector information can boost innovation and facilitate
economic growth. Vickery (2011) and others have concluded that “knowledge is a source of
competitive advantage in the ‘information economy’, and for this reason alone it is economically
important that public information is widely diffused” 22 .
Generating value from public sector information
Following a review of the evidence it is apparent that the two main ways in which value is being
generated currently, which will likely remain the case in the future, is through data discovery and
data exploitation. The former relates to making better and, in some cases more, public sector
information available and making it more accessible – what might be loosely term supply-side
considerations. The second dimension relates to using and re-using public sector information
better – demand-side considerations. This might be through reducing consumer risk aversion to
using public sector information, improving data exploitation techniques, changing cultures and
improving data analysis skills (which, in turn can improve competiveness).
Figure 4: Generating value from public sector information
Source: Deloitte analysis. Boxes not to scale.
Value is thus generated through exploiting existing datasets to identify insights through statistical
analysis, ‘data-mashing’ and visualisations. The value of public sector information in the UK can
therefore be increased both by increasing the quality (and quantity) of public sector information
(increasing the potential for data discovery) or by better using existing and new datasets.
While, in some cases, more public sector information being available can also increase value,
there is not a linear relationship between quantity of public sector information and its value. What is
21
In this case, accessibility refers to the conditions attached to the use and re-use of public sector information. These
conditions can take the form of a fee, limitations on who has access the data and limitations over how the data can be
used.
22
Quoted in Vickery, G (2011) Review of recent studies on PSI re-use and related market developments.
16
Market Assessment of Public Sector Information
key is the quality of this data and its amenability to data analytics. Simply releasing more datasets,
irrespective of their quality, is likely to only have a minimal value impact.
Value from different public sector information datasets
Clearly the value of public sector information will vary according to the identity of the final
beneficiary, the dataset in question and how and when the information is being exploited. Figure 5
highlights how different public sector information datasets can generate different levels of value.
Figure 5: Value of different public sector information datasets
It should be noted that in and of themselves, public sector information datasets do not carry any intrinsic value.
Value is a function of customers being able to use the datasets to generate revenue and jobs, assist in decision
making, and promote transparency and accountability – these impacts are discussed in the following sections.
The evidence points to the value of any given public sector information dataset to society as being positively
correlated with:

the content of the dataset – there are certain content themes that have well-established uses and re-uses
or may be fundamental to the provision of services, products and types of research. Where this is the case,
certain content themes (e.g. geo-spatial data or transport data) will positively influence the value of the
public sector information dataset to society;

the flexibility of the dataset – where datasets can be used in multiple ways to generate insights (e.g.
house price data that can be used as proxies for a range of factors such as environmental conditions or
school performance), the relative value of the dataset to society may be higher than a single-use dataset;

the accuracy, comprehensiveness and speed of refresh of the dataset – as one would expect, the
value of public sector information dataset will increase with its accuracy, comprehensiveness and speed of
which it is updated (e.g. economic statistics); and

the ability to link the data – the easier it is to link a given dataset with other datasets and other forms of
information will increase its flexibility and comprehensiveness, and again can increase its value to society.
Of course, it should be noted that the value of a given public sector information dataset to society will vary over
time, with some datasets becoming more or less valuable as new uses are discovered or rival datasets used –
and accordingly predicting the future value of any given dataset is difficult.
Building on existing analyses, it is possible to create a data intensity matrix to compare how different types of
datasets are used by different economic sectors in the UK currently (see Chapter 3 of the main report). The
analysis shows that the construction, real estate, financial and insurance, public sector and arts, entertainment
and recreation sectors are some of the largest users and re-users of open data 23 . Dataset types most
commonly being used include demographics, economic, environmental and geo-spatial and housing data.
From an analysis of data.gov.uk dataset requests, the most popular dataset category requested is location (or
geospatial) data. This reinforces the impression that this category of data generates some of the highest levels
of value. Environment, transportation, health and society are also data that attract particularly high levels of
requests. Requests for data used to scrutinise and hold government accountable, such as government, finance,
policy, administration and spending data are significantly lower, suggesting that this attracts less interest.
However, even in these categories the numbers of requests are not insignificant, and the threat of accountability
arising from the data, rather than the specific use of the data may be the primary driver of value.
Source: Deloitte analysis
Nonetheless, at a broad level, from the literature surveyed for this evidence review, it is clear that
the release of public sector information has the potential to generate significant economic value
through stimulating innovation, addressing market failures (such as barriers to entry and
information asymmetries), facilitating new ways of working and creating network effects
arising from more and more users of public sector information generating new insights and crossfertilisation of ideas, helping with the creation of new markets.
Conceptually, one can see how public sector information can drive economic growth and wider
prosperity in the form of happiness and sustainability. The simplified framework, shown below in
Figure 6, illustrates the long-term drivers of the UK’s economic growth and prosperity.
23
While this analysis has been carried out on the basis of open data, one might hypothesise that a similar picture holds
for public sector information.
17
Market Assessment of Public Sector Information
Measurable outputs (such as Gross Value Added 24 , employment and productivity) as well as less
easily measured outcomes (such as happiness and sustainability) are determined by two key
drivers in the form of productivity and employment levels. In other words, in the long-term and
considering the supply side of the economy only, the economic output of the UK is a function of
only two things: the number of people engaged in gainful employment and the amount each person
in employment is capable of producing.
In turn, these two key drivers of economic growth and prosperity are determined by seven
necessary enablers. These are related to the infrastructure required to facilitate long-term growth:
a deficit in these enablers will equate to a supply-side constraint on economic growth in the longrun. This could be caused either by limiting the growth in the working population (through
insufficient affordable housing capacity limiting labour mobility) or by acting as a drag on
productivity growth (through below-par ICT connectivity or a sclerotic transport system). The
amount each worker produces is determined by skill levels; the extent of innovation in products
and processes; the degree of investment in capital; entrepreneurial activity; and, lastly, levels of
competition. Expanding and enhancing these enablers can therefore positively impact employment
and productivity, which in turn can generate economic growth and greater prosperity.
Figure 6: Long-term UK economic growth framework
Source: Deloitte analysis based on HMT and BIS analysis. See www.hm-treasury.gov.uk/d/ACF1FBD.pdf for more
details.
As is shown in Figure 6, the introduction of more public sector information can have a positive
impact on infrastructure and other enablers. For example, it can stimulate innovation, help enhance
skills, promote competition and enterprise and attract investment to new products and services. In
terms of infrastructure, public sector information can help generate efficiency savings across public
services and improve business decision-making.
24
Which can be thought of as analogous to GDP.
18
Market Assessment of Public Sector Information
Understanding how public sector information creates value in the UK
Having reviewed the literature, it is useful to disaggregate the different types of value generated by
public sector information according to the beneficiary:

the direct value of public sector information to producers and suppliers (the PSIHs):
these are the benefits accruing to producers and suppliers of public sector information through
the sale of public sector information or related value-added services;

the indirect value of public sector information arising from its production and supply:
the benefits accruing up the supply chain to those organisations interacting with and supplying
PSIHs (but not directly using or re-using public sector information), and the benefits accruing to
those organisations where employees of PSIHs and supply chain organisations spend their
wages;

the direct use value of public sector information to consumers of public sector
information: the benefits accruing to businesses, civil society, individuals and the public sector
from directly using and re-using public sector information for a variety of purposes; and

the wider societal value arising from the use and re-use of public sector information: the
benefits to society of public sector information being exploited, which are not readily captured
elsewhere.
Using the above disaggregation, Figure 7 overleaf shows how value can be created across the
public sector information market, from the perspective of different participants along the supply
chain. Three (hypothetical) broad examples are shown:

a policy efficiency example where value is generated by the re-use of health data;

an app developer example where value is generated by the use of data as a key input into
transport apps that seek to save time; and

a data analytics example where value is generated by the use and re-use of data to generate
customer and business insights.
Value is generated across each element of the supply chain. PSIHs who produce and disseminate
the public sector information can create value through employing staff to collect and organise the
public sector information. If they also sell public sector information or value-added products, value
will be created through the revenue attributable to public sector information. Or, if the PSI is made
available for free under the Open Government License (OGL) this has the potential to maximise its
use by third parties to generate added-value in the supply chain (see downstream impact).
Upstream along the supply chain, value will be created as a variety of third party organisations
doing business with PSIHs (such as IT suppliers, operations and maintenance firms, caterers,
recruitment agencies, etc.) will receive orders from the PSIHs and earn revenue. Similarly,
businesses such as retailers where employees from PSIHs and other organisations in the supply
chain spend their wages will also earn revenue and generate jobs.
Downstream along the supply chain, value is created by consumers of public sector information
directly and indirectly using the information and data. In the policy efficiency example, this is
through using the public sector information to identify efficiency savings; in the app developer case
value is created through the sales of the app and the direct financial benefits to users in the form of
time saved; and in the data analytics example, value comes from efficiency savings and better
targeted products for consumers – raising revenue.
19
Market Assessment of Public Sector Information
Figure 7: Economic and social value from public sector information
Source: Deloitte analysis. In the Figure 7, the term Gross Value Added (GVA) is used to capture economic value (from profits, wages and rents).
Market Assessment of Public Sector Information
Even for these three archetypal examples, the wider societal value from public sector information is
more challenging to measure as it captures wider benefits arising from the use and re-use of public
sector information that are typically not measured in monetary terms. The literature discusses a
number of ways in which public sector information can have broader value impacts through:

increasing democratic participation: giving citizens and businesses access to public sector
information allows them to perform their own analyses of salient issues, make more informed
choices about public service providers and interact with policymakers to challenge their
assumptions and improve the policymaking process;

promoting greater accountability: for example through the scrutiny of costs of public service
provision and benchmarking comparable services;

greater social cohesion: for example, by providing more information on the provision and
distribution of services, public sector information can be used to dispel myths on who receives
certain public services;

generating environmental benefits: such as reducing congestion and pollution through the
release of better traffic and transport data which helps drivers to better plan journeys; and

identifying previously unknown links between different policy areas: through data-mash
ups it may be possible to develop system-wide solutions that holistically seek to address the
root of policy challenges.
This value is potentially significant and is likely to have a major influence on overall societal
wellbeing.
Estimates of the value of public sector information and value to PSIHs
Economic value
Figure 7 reflects the complex nature of the public sector information market and its participants for
just three very specific examples 25 . Very little evidence exists to accurately identify the ways in
which public sector information acts as an input in the productive process and the importance it
has in generating value. There is no central database tracking data and information collected and
stored by the public sector and many businesses are reluctant to disclose how they use and re-use
public sector information. As part of this research, steps have been taken to remedy these gaps
through stakeholder consultations, but the fact remains that it is necessary, in many cases, to
make a number of simplifying assumptions to arrive at indicative quantitative value estimates.
Whilst Deloitte is content that these estimates make best-use of the information provided within the
scope and parameters of the study, the estimates quoted in this report should be considered with
regard to these caveats.
Using an adapted bottom-up methodology outlined by the Office of Fair Trading in its 2006
Commercial Use of Public Information report 26 that is consistent with HMT Green Book guidance 27 ,
this report has estimated that the value of public sector information to consumers, businesses
and the public sector is currently approximately £1.8 billion in 2011/12 in 2011 prices (what is
termed the economic value). This figure includes the direct value of public sector information to
PSIHs, the indirect value of public sector information arising from its production and supply, and
the direct use value of public sector information to consumers of public sector information – for
shorthand, this is referred to as the narrow economic value of public sector information.
25
Figure 7 is also limited to the economic and social value within the UK. Clearly, there will be wider international
network and other effects that spillover outside the UK. These effects, which potentially could be significant, are outside
the scope of this report.
26
Available at: www.oft.gov.uk/OFTwork/publications/publication-categories/reports/consumer-protection/oft861. In
particular, see Annex G for the modelling framework. See Appendix 5 for a more detailed description of the methodology
used in this study, including assumptions around additionality and the counterfactual.
27
See www.hm-treasury.gov.uk/data_greenbook_index.htm
Market Assessment of Public Sector Information
Figure 8: Estimates of the current value of public sector information in the UK in
2011/12 (2011 prices)
Value category
Central scenario
High scenario
Low scenario
Direct consumer surplus
£1.6 billion
£2.00 billion
£1.00 billon
Producer surplus
£0.1 billion
£0.1 billion
£0.1 billion
Indirect and induced
value (supply chain and
consumer spending)
£0.1 billion
£0.1 billion
£0.1 billion
Total
£1.8 billion
£2.2 billion
£1.2 billion
Source: Deloitte analysis
Subjecting this modelling approach to sensitivity tests 28 generates a lower bound value of £1.2
billion, and an upper bound value of £2.2 billion for the value of public sector information in
2011/12.
Wider societal value
The use and re-use of public sector information has much larger downstream impacts affecting
all areas of society. Public sector information can act as a catalyst for positive creative
destruction – the process of generating innovation, identifying and making efficiency savings,
helping officials, business leaders and ordinary citizens make better policy and business decisions
and promoting accountability and transparency.
Through the use of case studies, it is possible to derive some indications of the financial scale of
different components of social value.
Figure 9: Case study examples of public sector information generating value
Sector
Example
Healthcare and life sciences

Using NHS prescribing data to identify efficiency savings of up to £1.4
billion per annum from switching from branded to generic drugs

Reducing mortality rates following cardiac surgery leading to a positive
impact on standards

Using traffic data provided by the Highways Agency to build a tool to
better plan road trips – potentially saving motorists up to £6.5 million per
annum (value of time saved)

Using public sector information on roadworks to better co-ordinate utilities
work and journey planning – leading to estimated benefits of around £25
million per annum for local authorities and road users in terms of
efficiency savings and reduced congestion

Embedding public sector information on real-time transport data in apps in
London can save users between £15 million and £58 million in terms of
time saved, each year
Transport
See main report for further details of report authors and details
Whilst these individual estimates of social value from the different case studies cannot be simply
summed (some are cost-savings, some are time-savings some are economic value and some are
non-quantifiable), they suggest that the total social value of public sector information is likely
to be significantly greater than the narrow economic value presented here. A conservative
28
Through altering assumptions on demand curve shapes, elasticities and the usage of public sector information
datasets. Alternating these assumptions does not materially affect producer surplus and indirect and induced impacts.
22
Market Assessment of Public Sector Information
estimate of the wider social value of public sector information suggests a value at least three times
the size of the £1.8 billion economic value may be appropriate 29 , i.e. over £5 billion p.a.
However, without further data on the relationship between public sector information use and social
outcomes, it is not possible to say definitively what the value might be.
Total value
Aggregating the calculated economic value and the conservative estimate of wider societal value
generates a total current value of public sector information of between £6.2 billion and £7.2 billion
in 2011/12 (2011 prices).
Future uses
Given the unpredictable nature of innovation, with the most valuable new products often being
among the least anticipated, it is risky to offer firm predictions of the future uses and value of public
sector information. Nonetheless, given the current trajectory of increasing volumes of public sector
information being made available in ever more accessible formats, it seems reasonable to
hypothesise that the larger part of the value to be generated from public sector information lies in
the future. Much of this value may come from combining public sector information with information
from other sources, such as details of consumer spending habits held by supermarkets and other
retail firms, or details of domestic and commercial properties.
Many of these future uses may be extensions of the ways in which public sector information is
already being used. For example, a recent study calculated that the benefit to the local government
sector from the use of geospatial data was £232m over five years, through a variety of measures
such as more efficient routes for waste collection. 30 It is to be expected that such uses will become
increasingly widespread throughout the public sector at both a local and central level, as more
information becomes available, methods of harnessing this information are improved, and best
practice becomes more widely adopted, thus increasing the value derived from public sector
information.
Further benefits are likely to emerge for private individuals, business and other organisations as
methods of exploiting public sector information are developed and improved. More effective ways
of data exploitation can include:

greater system-wide analysis through more sharing and combining datasets together to
consider policy and business issues from a range of non-traditional perspectives;

unlocking the potential from linked data: and better integrating; and

greater data-mashing with private sector and individual datasets.
The case studies explored in the main report illustrate some of the ways in which public sector
information is generating value. Such examples are likely to proliferate as available data
proliferates and the expertise to exploit it is developed. An example is the case of Honest
Buildings, described in Chapter 6: this SME is exploring ways to help businesses reduce their
energy bills by combining Energy Performance Certificate data with privately held data, potentially
delivering significant savings and reducing the UK’s energy usage. Such examples may be
expected to proliferate as access to information increases, and innovative uses are developed.
In addition, there are likely to be benefits from wholly new uses of public sector information.
Numerous additional possibilities for achieving such savings and improving policy may become
apparent as more information is made available and expertise in its use increases. In addition,
there are likely to be more ‘blue sky’ opportunities, such as those identified by the Government
29
This is based on insights from individual case studies by other authors which have variously estimated the wider value
from public sector information use in the health care sector, geo-spatial, meteorological and other sectors in the order of
billions.
30
The Value of Geospatial Information to Local Public Service Delivery in England and Wales, 2010
23
Market Assessment of Public Sector Information
Office for Science. These are explored further in Chapter 5, but include improving management
and resilience of food supplies, tackling obesity, and detecting and identifying infectious diseases.
These require combining public sector information with other sources of information rapidly and
making the resulting insights available to decision makers in a timely fashion to allow prompt
responses to crises and informed policy decisions.
Barriers to realising the full potential of public sector
information in the UK
A taxonomy of barriers
Adapting a taxonomy originally developed by the Data Strategy Board, the research has examined
the evidence to identify the areas where barriers may currently exist. Note, some of the themes
under each barrier overlap, e.g. privacy could be captured under legislation or access.
Figure 10: Taxonomy of barriers
Source: adapted from DSB
Legislative barriers include whether certain acts of legislation and other regulations reduce the
usability of public sector information datasets by consumers. There are also further questions over
whether current assurance/accreditation standards for public sector information remain effective
and whether there is scope for existing licensing arrangements to be improved.
Economic barriers include questions of which datasets to release (which datasets yield the
highest value) and how their costs (if any) should be covered.
Access barriers to maximising the value from public sector information can include:

a reluctance to publish or share datasets;

a bias against using non-traditional datasets;

a reluctance to use public sector datasets because of concerns around quality, reliability and
on-going support; and

a lack of skills to fully exploit the value of public sector information.
It should be noted that the extent of these barriers may differ between public sector information
datasets and also between the different types of users.
24
Market Assessment of Public Sector Information
Each of these barriers is discussed in more detail below.
Legislative barriers
Privacy and the impact of current regulations and legislation
The review of available evidence suggests that by and large, the current legislative and regulatory
environment around public sector information is not acting as a barrier to generating value and
market development. While a full legal analysis has been beyond the scope of this study, particular
acts and regulations such as those covering Data Protection and Human Rights legislation
have not emerged as preventing the development of new products or services using public
sector information 31 . The majority of stakeholders consulted have not reported that these generic
legislative acts are currently preventing them from using and re-using public sector information,
but, in some cases specific legislation controlling a particular data collection exercise do.
However, the evidence received does suggest there are some current challenges around
perceptions and attitudes towards data release. In particular:

regulations such as Data Protection are sometimes used as a shorthand justification for
not sharing public sector information within the public sector, with PSIHs not always able
to translate their awareness of their rights and duties into scenarios where public sector
information is released or shared, causing a barrier; and

when a policy decision is taken not to release public sector information datasets to the
general public, the reasons are often not well articulated or the conditions attached to
access are overly restrictive.
Stakeholders have cited examples where the Data Protection Act (and other legislation) has been
used as a reason for a PSIH to withhold information from the general public. However, as noted
above, in reality, the Act should rarely be a barrier to sharing information. Where public sector
information datasets do contain personal details, these may need to be aggregated or anonymised.
Where this is done effectively the Data Protection Act, in most cases, no longer applies to the
information and it can be made available.
This is not to downplay data protection issues. The impact of breaches leading to the release of
personal information can be extremely serious. Further, even if data is anonymised, people may be
reluctant to report it if they believe it can have wider consequences, e.g. reporting crime data may
adversely affect house prices 32 . These issues notwithstanding, PSIHs are beginning to use
innovative methods to test different ways of anonymising datasets – for example the Ministry of
Justice worked with statisticians, the private sector and the academic community to avoid the
‘jigsaw effect’ 33 occurring from the release of offender data.
Within the public sector, there are also questions surrounding the ability of different public bodies
to share datasets with one another. These have been explored in depth in the recent report by the
Administrative Data Taskforce 34 . It notes that as regards the legal gateways established to allow
departments and other public sector bodies to share information without obtaining the consent of
the data subjects, “recent experience demonstrates that link-specific gateway legislation is both
cumbersome and inefficient.” As a solution to this, the Taskforce recommends the creation of a
generic legal gateway to reduce this barrier to the exploitation of public sector information and
clarify the legal position around sharing data. Note that this applies to sharing of data within the
31
This observation should be caveated with the note that, at the time of writing, there are a number of initiatives to revise
various regulatory frameworks at the European level.
32
See for example: www.guardian.co.uk/money/2011/feb/01/police-crime-website-house-prices
33
The process of combining anonymised data with auxiliary data in order to reconstruct identifiers linking data to the
individual it relates to.
34
Source: Administrative Data Taskforce (2012), www.esrc.ac.uk/funding-and-guidance/collaboration/collaborativeinitiatives/Administrative-Data-Taskforce.aspx
25
Market Assessment of Public Sector Information
public sector and with certain accredited third parties such as researchers, rather than publication
for use by the general public.
Insights
 Existing regulations and legislation governing public sector information do not appear
to be acting as actual barriers to realising the full potential value of public sector
information. However, the manner in which some of these regulations and legislation
are interpreted can lead to overly risk averse behaviour and can create barriers.
 Increasingly effective anonymisation techniques and an approach across the public
sector that emphasises granting access to as much data that is compatible with
privacy and security has the potential to improve access to public sector information.
Licences
Licensing conditions play an important role in facilitating (or preventing) the full exploitation of the
value of public sector information. The ideal standard widely acknowledged by stakeholders was
licensing public sector information under the Open Government Licence (OGL), and increasing
amounts of public sector information are being made available as open data under it, even by
organisations that have traditionally charged for data 35 . Some groups of stakeholders have argued
that if the OGL were used for all public sector information, this could substantially increase the
openness of the UK’s public sector information and could remove many of the barriers to use and
re-use that currently inhibit the realisation of the full value of public sector information.
However, some argue that the public sector information landscape is characterised by its diversity,
and for this reason a ‘one size fits all’ solution is unlikely to be practical or necessarily desirable.
For cases where the release of information under the OGL is not considered appropriate and cost
recovery is justified, the introduction of a generic charging licence could, in principle, address the
current complexity of charging arrangements for public sector information, completing the UK
Government Licensing Framework and simplifying licensing arrangements for users. Some
stakeholders pointed out that by having a charge, users and re-users can be re-assured of the
quality of data and be able to expect a certain level of service (this is discussed in more detail
below). However, one concern raised was that such a licence could encourage some PSIHs who
are not covered by the OGL, to charge for certain datasets.
Insights:

Making public sector information available under the Open Government Licence is
seen by a large number of stakeholders as an effective means of removing barriers to
exploiting the value of public sector information.

If there are instances where a generic Government Licence for charging for public
sector information are applicable, such a Licence would need to be drafted in a way to
avoid incentivising charging when there is no strong justification or leading it to
become the default alternative to the Open Government Licence.
Eligibility restrictions on accessing particular datasets
With respect to access restrictions to public sector information datasets (such as the datasets only
being available to researchers or in secure environments), in many cases the rationale for these
restrictions are clear. The rationale may cover national security reasons or genuine data protection
concerns. However, some stakeholders have reported that in some cases, the rationale for
restrictive access to certain datasets is not always clear, appears overly restrictive or may
35
For example, Ordnance Survey has released 11 datasets, and the Met Office, Land Registry and Companies House
have each opened up some of their data.
26
Market Assessment of Public Sector Information
no longer be relevant, although it should be noted that just because a stakeholder feels the
rationale is overly restrictive, that does not necessarily make it so.
For example, some commercial start-up companies have reported difficulties in accessing different
versions of the National Pupil Database as non-research organisations. While restrictions on
access are in this case clearly needed to protect sensitive and personally identifiable information;
there is a valid question as to whether there are ways that some or all of this information could be
made available to commercial organisations to develop new products and services to help parents,
teachers and other stakeholders to improve decision-making, increase accountability and identify
good practice. Without this, opportunities for innovations using this and other datasets are likely to
be missed, meaning loss of the value that they could potentially create.
Insights:

The eligibility restrictions imposed around certain datasets are typically due to
reasons of national security, data protection and other sensitivities.

The rationale behind these restrictions is not always clearly articulated and may, in
some cases, no longer apply. In some cases, this may be preventing opportunities for
innovation to take place.
Economic barriers
The evidence reviewed suggests a number of key economic challenges around maximising the
value from public sector information in the UK. Primarily these relate to the value and cost of
datasets, which can be grouped into two key themes:

which datasets are made available to the general public; and

which datasets consumers have to pay to access for (funding models).
Data gaps
There is currently no national or local information asset register covering the amount and type of
public sector information datasets held 36 (irrespective of whether it is being released to the public)
or the detailed costs of collection and dissemination. Similarly, there is little evidence on how
consumers themselves are using and re-using particular datasets.
While this has meant this report has been forced to make a number of assumptions in order to
reach quantitative estimates, there is a more significant impact on PSIHs themselves. By not being
able to accurately ascribe value to different datasets, PSIHs are generally unable to reach
evidence-based decisions as to which datasets to publish, how to publish them and what support
to provide. One consequence of this is that the costs (which are readily measurable a priori) of
making particular public sector information datasets available may appear to be much larger than
the benefits – as the benefits case cannot be clearly articulated due to a lack of evidence.
Indeed, as part of this report, a sample of government officials have been consulted over the
reasons why public sector information datasets are not released: over 50 per cent of those
responding identified resources as an issue preventing the release of more public sector
information, with the following response being typical: “in a time of diminishing resources and the
need to make best use of the resources we have; the time and the cost of ensuring data validity
before release is an issue.” This was one of the most common reasons given for why data is not
released, but not the only one – others include data protection and legislation.
Stakeholders have identified particular datasets (such as an aggregate Energy Performance
Certificate database) that are currently not available to them (either for free or for a fee) that they
could use to generate new products and services. By not making these datasets more widely
36
Though the UK is not alone in this respect.
27
Market Assessment of Public Sector Information
available (or articulating the reasons they are not available), PSIHs may, whether knowingly or not,
be acting as a barrier to realising the full potential of public sector information.
There are routes to address these data gaps. One route might involve conducting a public sectorwide audit of public sector information (which also covers the customers using and re-using it).
This would involve both primary research to determine the stock of public sector information held
across all PSIHs and also identifying how this data is being used commercially across different
economic sectors and actors. This would help match costs and benefits to individual datasets and
building this report to identify those datasets that can generate the most value for the UK. The
costs of doing such an audit would not be insignificant and for it to be effective, it would need to be
repeated in future years (though the on-going costs may be lower than the one-off start-up costs).
However, the benefits of doing such an audit could be outweighed by PSIHs having a much better
sense of which datasets are creating the most value, which can guide future policy decisions
around releasing data.
Other routes to addressing data gaps might involve better tracking of usages (perhaps through
measuring how often data is re-used in other sources), requiring all new public sector information
datasets to be logged centrally or crowdsourcing an audit across the public sector. Another
potential route for the long-term, at least conceptually, might be a true single conduit for the access
of all public sector information.
The Open Data User Group (ODUG) has been set up as an advisory group to Government and
provides a channel through which the potential users of data, from all areas of the community, can
identify their need for data and set out the benefits they expect this data to deliver. This will help
PSIHs understand in more detail the potential benefits which the release of individual datasets can
be expected to deliver.
Through the completion of data gaps, it would be possible for government to articulate what is
meant by the term core reference dataset (beyond the definition contained in the Open Data
White Paper). The issue of core reference datasets is also raised in recent work by APPSI on the
national information framework for public sector information and open data. 37 Consideration could
include:

the features of a particular dataset that make it a core reference dataset;

the funding arrangements of core reference datasets (see below);

the obligations of suppliers of core reference datasets; and

the rights of citizens to core reference datasets.
Insights:

There are significant data gaps when it comes to public sector information. This lack
of data can, in some cases, lead to inertia with certain public sector information
datasets not be released or conversely, undue attention being given to datasets that
are unlikely to generate significant value but have a low cost of dissemination.

There are a number of routes to addressing these data gaps. These range from a
detailed, regular audit of public sector information to improved tracking of current
usage and a single conduit for all public sector information.
37
The issue of core reference datasets is also raised in recent work by APPSI on the national information framework for
public sector information and open data, www.nationalarchives.gov.uk/documents/nif-and-open-data.pdf.
28
Market Assessment of Public Sector Information
Funding models
Related to the question of making the business case for more public sector information being
released is the issue of how the cost of generation, collection, retention and dissemination of the
datasets should be funded. This also leads to questions around pricing for access to public sector
information.
This is a complex issue and data on costs (and benefits) is not fully available. Further, one must
take account of different costs being incurred across the different stages of the data lifecycle. For
example, in the case of so-called ‘exhaust’ or ‘by-product’ data that is generated as a result of
PSIHs conducting their day-to-day and other activities and duties, the marginal cost of its
generation will, by definition, be negligible or very low compared to the activity that caused the data
to come about; but the marginal cost of its dissemination to the wider public may be higher and
vice versa for some ‘purposely collected data’. There will also be additional costs in formatting the
data for use and further costs if support is provide to the public in re-using the data.
The issue of charging and funding models for public sector information is highly contentious and
extremely complex. For example, there remains an element of subjectivity as to what constitutes a
dataset and what constitutes a ‘value-add’ service – with some PSIHs arguing that what is being
charged for is not the public sector information itself, rather its interpretation and analysis.
As part of this report, a wide range of opinions have been expressed as to whether there is any
economic or other justification for charging for public sector information datasets. On the one hand,
these arguments include:

charges for datasets create barriers to entry and expansion for SMEs and individuals to
develop new products and services;

the charges prevent SMEs and individuals from ‘experimenting’ with the datasets before
they purchase to see if they are able to derive value from them, thereby making it hard to
develop business cases; and

any lost revenues to PSIHs from releasing datasets for no cost will be recovered by the
Exchequer in the long-run through increased tax revenues and more jobs being created.
In contrast, arguments have been put forward to support current pricing arrangements include:

aligning a revenue stream with a particular dataset will ‘protect’ it from any reductions in
funding, allowing PSIHs to continue to supply this even if they themselves must make other
savings;

a price can be interpreted as a signal of consumers’ willingness to pay for a particular
dataset’s quality and a commitment by the PSIH to maintain this and offer support; and

charging for certain datasets is necessary given they include elements of commercial or
international datasets.
As the most visible PSIHs that charge for certain public sector information, a great deal of attention
has been focused on the four Trading Funds that make up the Public Data Group (PDG):
Ordnance Survey, the Met Office, Land Registry and Companies House. The view has been
expressed by some stakeholders that these PSIHs hold some of the most valuable datasets and
there are strong arguments that these should be treated as core reference datasets available to
all at no direct cost to the general public.
It is not helpful to treat these PSIHs as a single group – indeed, there are a number of other public
bodies that charge for access to datasets 38 and there are a number of other Trading Funds outside
the PDG. The four Trading Funds in the PDG differ in their sources of revenue, their public service
duties and whether the public sector information they hold can be classed as ‘exhaust’ data or
38
For example, the Office of National Statistics has a subscription charge for certain datasets.
29
Market Assessment of Public Sector Information
‘purpose-collected’ data. Further, it should be noted that substantial progress has been made by
the four Trading Funds in recent years to make increasing volumes of data available as open
data 39 and there continue to be moves in this direction.
However, despite these positive steps, there remains a perception among many consumers
and commentators 40 that they are unable to access certain datasets for reasons of cost and
this is creating a barrier to business growth. A number of studies (e.g. Pollock, 2011 and others –
see Appendix 3) have argued that releasing these datasets as open data will have significant
welfare benefits.
The impact of a cost recovery model for public sector information, compared to the Open
Government Licence model, is summarised below.
Figure 11: Potential impact of the cost recovery model on the public sector
information market
Taxpayer funded public sector information
Public sector information supplied on a cost
recovery basis

Public sector information typically made
available at no cost to users under the Open
Government Licence

Access to public sector information involves a cost to
the user and may come with restrictions

There may be strong public policy reasons for
having these datasets available at no cost

The costs of collection (rather than dissemination)
are significant and cannot be borne by the public
purse

Will not include other commercial or international
datasets

May include other data sourced under licence

The availability of public sector information at no
cost to the user can contribute to the
transparency agenda by increasing access to
the widest possible customer base, irrespective
of the ability to pay

Access to Public sector information is restricted on
the basis of cost

Will not typically include other services or any
other guarantees

May also include bespoke value-add services and
guarantees of data quality and continuity
Source: Deloitte analysis
It is very difficult to perform a robust cost-benefit analysis of different funding model options for
PSIHs, not least because PSIHs differ greatly from one to the next. While some data is available
as to the costs incurred from collection and dissemination, this is not typically openly available or
apportioned by dataset, nor is data readily available on how their customers are using the datasets
and generating revenue and value.
More fundamentally, very little data is available on what the benefits might be if charging models
were to radically change – as it is very difficult to predict how businesses and individuals might use
datasets in the future to generate new products and services and by implication impact economic
growth. It is also important to note that, as per the HMT Green Book 41 guidance, any benefits from
a change in charging structures should include not just increased tax receipts but wider social
benefits and costs in terms of organisational impacts.
Even in the case of the four Trading Funds that make up the PDG, estimating the effects on
Exchequer revenue of releasing all their public sector information as open data is a difficult task,
even in spite of the information made available by the Trading Funds as part of the study. There
39
For example, see www.ordnancesurvey.co.uk/oswebsite/products/os-opendata.html
See for example www.freeourdata.org.uk/ and other references in Appendix 3.
41
Available at www.hm-treasury.gov.uk/data_greenbook_index.htm
40
30
Market Assessment of Public Sector Information
are differing views on precisely which revenues should be considered relevant to PSI, considering
factors such as specific revenues from the data, the cost of collecting the data, the extent to which
Government ‘buys’ data from itself and so forth.
That notwithstanding, on the basis of the information made available to this study, it is possible to
indicatively calculate that the cost effects on Exchequer revenue of continuing to collect and
disseminate Trading Funds’ public sector information in its current guise without charging for it.
This cost is estimated to be in the order of £395 million on an annual basis 42 . However this figure is
without regard to the extent that government pays for public sector information from the four
Trading Funds in the PDG (this varies significantly between Trading Funds). On the basis of
information provided by sales channel, it can be estimated that the annual loss to the Exchequer
would be lower, as government would no longer need to purchase these public sector information
datasets – it could use them at no cost. In this scenario, the direct loss to the Exchequer on an
annual basis might be of the order of £143 million. This figure may be lower still if there are
efficiency savings to be made if fewer dedicated sales and marketing resources are required by
Trading Funds.
Following the HMT Green Book approach to account for the wider social and economic benefits, it
is important to note that in a world without charging, private sector entities (consumers and
businesses) that currently pay for access to public sector information provided by the Trading
Funds would benefit by this amount – £143 million – and some of this may be recouped in the
medium-term as a result of additional economic activity generating tax revenues.
An additional group that is currently deterred by having to pay for the data would also benefit as
they are able to access the data. Estimating the size of this latter group is difficult but directionally it
is clear that removing charging would mean more people and businesses, not fewer, would be able
to access and benefit from this data 43 . Conversely, organisations who are at an advantage in using
their own proprietary information for commercial advantage, might find their competitive advantage
eroded if more public sector information is released to act as a publically available substitute to that
data.
The situation is thus complex, and while the above example is stylised, it does suggest the
quantum necessary for the associated benefits to outweigh the costs . Without more detailed and
accurate data on both costs and benefits (including wider social benefits) it is not possible
to reach a clear conclusion on this issue.
However, this is not to say there is no room for improvement today in the provision of public sector
information that carries a charge and a number of steps are being taken in this respect. These
include 44 :

much more communication about what existing licences allow consumers to access and use /
re-use the public sector information for, building on existing efforts by the trading funds and
other PSIHs to build awareness among the user community;

offering substantial discounts to SMEs and individuals;

implementation of a ‘royalties’ model for consumers to exploit the value of public sector
information up to a certain value before a charge is applied;

greater provision of ‘sandbox’ or secure environments in specialised locations across the UK
to allow consumers to explore datasets;

greater provision of out-of-date public sector information at no cost to the general public to
allow consumers to experiment; and
42
43
44
This figure comprises of Trading Funds’ operating surpluses and the cost of data collection.
There would accordingly be second-round effects for people consuming products provided by these businesses, etc.
These options have not been costed.
31
Market Assessment of Public Sector Information

greater use of ‘hack days’ to demonstrate the value of particular public sector information
datasets.
Of course, not all of these initiatives will be applicable to all PSIHs that charge for public sector
information – there is no ‘one-size-fits-all’ model – but they should work with the grain of the market
and build on existing initiatives to release more data. In some cases there may be significant
logistical or legislative challenges to overcome before the above suggestions are implemented.
Insights:

The issue of charging for public sector information datasets and their funding models
is complex. While significant progress has been made by a number of major PSIHs in
this area in simplifying charging structures and making more public sector
information available as open data, there remains a perception that barriers exist and
there is scope for improvement.

Ways to improve access could include greater communication on licence conditions,
discounts to certain consumers, a royalties model, use of a ‘sandbox’ model, greater
provision of out-of-date information and more hack days.

There is currently a lack of data to definitively conduct a cost-benefit analysis across
different funding models across the range of PSIHs currently charging. This is an area
for further analysis through primary research with the direct customers of public
sector information and a detailed cost apportionment exercise to assign costs to
individual datasets.
Access barriers
The evidence reviewed suggests a number of access barriers to fully maximising the value of
public sector information:

difficulties around finding where public sector information is located;

a lack of skills and understanding to fully exploit public sector information;

the format and reliability of public sector information; and

a reluctance to use and share public sector information.
Public sector information fragmentation
Fragmentation in the supply of public sector information continues to be a problem for many
consumers, even following the establishment of a number of data portals. A sample review of
websites done as part of this study has found that too often it is difficult to locate a clear point of
contact, or establish who has ownership of and responsibility for a particular dataset. Even on
data.gov.uk, where contact details are provided, these are often generic enquires email addresses
rather than named contact email addresses.
In contrast, this report has received feedback that when consumers have found the relevant point
of contact in a PSIH their experience has been very positive with a productive dialogue being
established. Indeed, some individual PSIHs have established clear procedures for consumers
seeking to raise queries or challenge restrictions, although this is not yet happening in all areas of
the public sector. Dialogue can also take place between users: data portals such as data.gov.uk
and the London Datastore have busy forums and request pages which encourage an active
dialogue between information users and re-users and information holders.
Public sector information fragmentation can, at best, raise transaction costs from dealing with
public sector information and, at worst, deter users and re-users from using public sector
information altogether, or make this impossible to achieve. Reducing fragmentation across the
32
Market Assessment of Public Sector Information
public sector could save up over £50 million per annum in terms of reduced transaction
costs and time saved for data specialists 45 .
However, it should be noted that the existing fragmentation and opacity of some public sector
information datasets has created market opportunities for some intermediaries to develop
new products that aggregate datasets and present it in innovative ways. Thus, PSIHs need to
consider who is using their data and how, and whether there is a risk of crowding out innovation if
they intervene to reduce fragmentation in such a way that duplicates what the private sector can
provide.
Insights:

Improvements are ongoing to reduce the time taken to locate public sector
information datasets. Reducing this can help reduce transaction costs and the overall
cost of doing business.
Data scientists and the skills gap
A number of studies have recently been published contending that advanced economies face a
skills gaps in so-called ‘data scientists’. Although statisticians and experts on quantitative analysis
have long existed, data scientists differ from these existing professions in a number of important
ways. As well as being able to work with large volumes of structured and unstructured data, they
are able to translate these analyses into policy and commercial-ready insights and effectively
communicate them to a range of stakeholders, often using innovative tools and visualisations. Key
to this is the ability to “identify rich data sources, join them with other, potentially incomplete data
sources, and clean the resulting set.” Data scientists will also be conversant with the vocabulary of
public sector information and open data and have the skills to create and manipulate large
datasets and linked data. In many ways they therefore resemble scientists more closely than
traditional data analysts.
There is a fear that a lack of data scientists will reduce the UK’s competitive advantage. The
evidence received suggests that in the UK:

there is increasing demand for individuals with a portfolio of skills able to manipulate
quantitative data, present it in innovative ways and generate commercial and policy insights
from it;

many of the individuals performing these roles have no specialised training, but rather have
learned on the job and / or have a science/computation/mathematics background;

businesses rarely designate specific ‘data scientist’ roles; rather, such analyses are done
across a combination of professions such as statisticians, economists, researchers, analysts,
policy and commercial managers – a dedicated data scientist would embody elements of all
these roles;

certain industries such as pharmaceuticals, financial services, professional services and retail
are increasingly dependent on these skills sets and a shortage of them would reduce the UK’s
international competitive advantage.
Based on ONS figures for 2011, around 1.5 million workers in the UK, representing around 5
per cent of the active workforce, are employed in job categories that are likely to involve
elements of the role of the data scientist, but which individually may not be termed data
scientists. The average annual median wage of these workers was over £36,000 in 2011 which is
higher to the national annual median wage of £26,000.
45
Based on an assigned average value of time spent by individuals in occupations that regularly use public sector
information. See main report and appendices for details of calculations.
33
Market Assessment of Public Sector Information
Economic theory suggests that a skills shortage will manifest itself in the form of large wage
differentials between ‘data scientists’ and other comparable professionals. A recent report found an
observable pay premium for ‘big data’ staff in 2012, with salaries around 20 per cent higher than
those for IT staff as a whole. 46 This may persist due to lags between training and entering or reentering the workplace, but economic theory suggests it will eventually dissipate in the long-run as
supply increases to meet demand and is able to exploit the full value of public sector information.
While evidence 47 suggests there are a number of sector initiatives to reduce the talent learning
curve and establish precedence, these initiatives will take time to work through the system with the
interim consequence being that some value of public sector information may remain locked
up. The general scarcity and increasing competition for these skilled workers from the private
sector makes it harder to construct the infrastructure for world class public sector information. A
shortage of data scientists also hinders efforts to scale-up public sector information data analytics.
Within the public sector, concerns have been highlighted over a lack of skills and familiarity to
work effectively with data. These concerns should not be overstated as public sector officials
have a long history of using public sector information to inform policymaking without having
dedicated data scientists. What the concerns appear to be directed at are cultural biases against
using public sector information from outside home PSIHs, as well as having the necessary skills to
combine and manipulate Big Data and Linked Data. One stakeholder has observed that “data
[owners] do not fully appreciate the power and potential of open data. There may also be cultural
resistance to change: for instance, through not trusting or being able to exploit new, untried and
untested third-party datasets in analysis and policy advice.”
Where public sector employees do not have a numerate background, or are not accustomed to
working with large datasets, they could benefit from increased training and support in this area.
However, this research reveals understandable concerns over resources at a time when budgets
are under considerable pressure. The main report outlines a number of cost-effective routes to
build the public sector skills base, which include taking massively open online courses (MOOCS),
creating incentives to use public sector information as part of day-to-day business, and convening
the public and private sectors to work together to explore public sector information.
Insights:

While there may be gaps in the supply of ‘data scientists’, economic theory suggests
that in the medium- to long-term, the number of data scientists will increase, filling the
supply gap and reducing the current wage premium.

However, in the short-term, this may mean public sector information is left underexploited and associated value remains locked out. The general scarcity and
increasing competition for these skilled workers can make it harder to construct the
infrastructure for world class public sector information and scale up efforts to exploit
its value.

There are some low-cost solutions that can be explored in order to quickly improve
the skills base to be able to effectively manipulate and extract value from public sector
information.
Format and reliability
The release of data in an unfinished form raises concerns for the public sector too because of
concerns that the public may be provided with data that is inaccurate and potentially misleading,
which can have negative consequences.
46
See e-skills UK, ‘Big Data Analytics: an assessment of demand for labour and skills, 2012-2017’ (January 2013)
See Deloitte Tech Trends 2013, available at
www.deloitte.com/view/en_GB/uk/services/consulting/technology/technologytrends/abbffbfdad4ac310VgnVCM3000003456f70aRCRD.htm
47
34
Market Assessment of Public Sector Information
With respect to the format of public sector information, its importance may vary between
customers. For casual consumers, it is clearly of the utmost importance, and they may most value
format. In contrast, professional users may be more concerned with consistency of service and
data and can accommodate changes to format. Equally, intermediaries may positively value low
quality data as they can provide a service to improve the quality and format of public sector
information for wider consumption.
Evidence as to how significant a barrier this currently forms to the use and re-use of public sector
information is mixed. In general, there seems to be steady improvement in all the areas of format,
although there remains work to be done – especially with regard to ensuring consistency in format
and upgrading the star rating of datasets. Progress is also being made with data being updated
more consistently. Examples of this include data released by the Met Office, and transport data
released directly by Transport for London as well as through data.gov.uk. The Open Standards
Principles 48 , the Standards Hub 49 , the Open Standards Board and the public sector will be
crowdsourcing, researching and implementing data standards for Government IT systems. The
user challenges it will focus on are likely to cover standards relating to formats and meta-data that
should help to provide data on reliability. Some of these open data standards may be made
compulsory for central government use.
However, a minority of stakeholders have complained that datasets released by different PSIHs
are not always easy to combine and work across, because of variations either in the content or the
format. This is a particular issue with data produced by local authorities, with each local authority
often adopting their own standards and procedures. This report recognises that there are on-going
efforts, such as e-PIMS, to tackle this issue and secure a greater degree of standardisation across
PSIHs.
Further, there appears to be scope for improvement is greater certainty and clarity over the
publication schedule of public sector information datasets and what users can expect from PSIHS.
In particular, stakeholders have noted the value in having a cover sheet setting out the limitations
of the data, its release schedule, explaining outliers and providing links to previous analyses (i.e.
greater use of metadata 50 ).
Insights:

Improvements continue to be made on the quality, format and consistency of public
sector information.

There is scope for improvement, especially around the greater provision of metadata.
Summary of key research questions answered
Further details behind each of the key findings, addressing the Data Strategy Board’s overall
research questions are summarised in Figure 12 below.
48
See www.gov.uk/government/news/government-bodies-must-comply-with-open-standards-principles
See http://standards.data.gov.uk/
50
That is, broadly speaking, data about data.
49
35
Market Assessment of Public Sector Information
Figure 12: Key research questions answered
Theme
Findings
Definitions of public sector information and its characteristics

What are the main market segments for public sector
information both existing and emerging?

While there is no central figure on the total number of public sector information datasets currently
being made available across all parts of the public sector, a review of selected data portals suggests
the number could exceed 37,500 from over 750 different publishers with over 2.5 million downloads

What is the composition of these markets in terms of
companies, their governance and their outputs

The key suppliers of public sector information are public sector bodies. An initial assessment shows
that of a total of 781 publishers (as of 27th February 2013 on selected data portals), the top ten
suppliers of PSI supplied over half of all datasets published on the website, with the Office of National
Statistics, the Department for Communities and Local Government and Health and Social Care
Information Centre the three largest suppliers by number of datasets.

Customers of public sector information including the private sector, civil society, individuals and
government itself. Business models include developing apps that use/re-use public sector information,
data enrichers, data enablers and data marketplaces.

Analysis suggests the construction; real estate; financial and insurance; public sector; and arts,
entertainment and recreation sectors are some of the largest users of public sector information (and
public sector open data).

As well as being used for commercial and policy purposes, public sector information can be used for
scientific research, data journalism and holding public service providers to account.

Based on ONS figures for 2011, it is estimated that around 1.5 million workers in the UK (5 per cent of
the active workforce) are employed in jobs that are likely to involve direct exposure to public sector
information and big data. Analysis of ONS data of jobs that involve ‘data science’ suggests an average
annual median wage of over £36,000 compared to the national annual median wage of £26,000 in
2011.

The extent to which public sector information is a key input into companies’ products and services will
depend on the company in question, with smaller companies likely to be more reliant on it than larger
companies, with their products and services more likely to use public sector information as the critical
input.

Public sector information acts as an input as either the main or supplementary data point for complex
algorithms and analyses used in products and services, as a source of insights (perhaps summarised
as data dashboards or visualisations) for business or policy decisions, as API feeds for apps and so
forth.

The supply side of the market is mostly populated with public bodies supplying public sector
information – there are few private sector substitutes.

What are the characteristics of the markets for public
sector information, including secondary markets?
Market Assessment of Public Sector Information
Theme
Findings

Broadly speaking, public sector information customers on the demand side of the market can be split
into seven archetypes: larger data companies, SMEs creating apps, SMEs creating efficiency
solutions, not-for-profit organisations, individuals, the public sector and other non-data specialist
companies. These companies vary in size (from sole traders to larger multi-national companies),
turnover (those having established, profitable businesses and those who have yet to report a profit),
the dependency on public sector information and the types of public sector information used.

The demand side of the market for public sector information includes a range of different business
models including: data marketplaces/infomediaries, apps-based re-use, data enrichment and data
enablers.

Downstream, insights and inputs from public sector information can be found in a range of products
and services. Products and services range from credit scoring, economic forecasts, weather apps,
transport apps, navigation systems and other mapping tools, research reports, tools to assist choice
(e.g. in education, health or housing) etc. Downstream, public sector information can also be used to
improve transparency and improve public services and hold providers to account.

What do past studies tell us about the markets and
how they function?

There is very little previous literature on the functioning of public sector information markets, possibly
due to data limitations around the supply of datasets. The literature has instead focussed on
delineating links between public sector information and growth and welfare.

What does the current evidence tell us about the
likely evolution of these markets?

The available evidence suggests that as better quality public sector information becomes available
and accessible, customers will be able to use it in increasingly sophisticated ways to exploit its value
and drive growth outcomes.

The literature suggests the provision of more and better public sector information can raise sales (a
5.7 per cent increase in sales in restaurants displaying ‘good’ scores), improvements in public health,
a productivity boost (around 0.6 per cent of GDP in the case of New Zealand geo-spatial data) and
faster growth (firms growing 15 per cent faster in countries where certain data was either free or priced
at marginal cost) – see Appendix 3 for more details. These figures are based on particular examples
and are not directly transferable to the UK.

Adapting the McKinsey analysis used by Policy Exchange, improvements in efficiency between 1 and
5 per cent caused by better use of public sector information can lead to annual savings of between £1
and 8 billion nationally and around £70 million in local government (on a smaller savings ratio).

Benefits are likely to come from further release of more and better data, but the release of swathes of
data in and of itself will not guarantee value, it is how society as a whole innovates with it that counts.
How public sector information is used and re-used inside and outside of government

What are the different types of public sector
information and can we estimate the value that they

Public sector information differs by its availability to the public, its format, its complexity and speed of
replacement, its content, its channel of distribution and the costs of collection, generation,
37
Market Assessment of Public Sector Information
Theme
Findings
hold for government and the economy?


What does the evidence tell us about the size and
potential market for public sector information? Is
there reliable data available?
What are the gaps in the evidence base and what
can we do in the short term to address them?
maintenance and dissemination.

This report has estimated that the narrow value of public sector information to consumers, businesses
and the public sector in 2011/12 was approximately £1.8 billion (2011 prices). This is a mid-point
estimate, with the sensitivity analyses giving a range between £1.2 billion and £2.2 billion

However, the use and re-use of public sector information has much larger downstream impacts
affecting all areas of society beyond its direct customers. Case studies examining the downstream
impacts created by different organisations using and re-using public sector information have shown
the monetary value of some of these benefits can be in the order of millions, if not billions.

For example, publishing data on adult cardiac surgery has estimated to reduce mortality rates, which
in turn has an economic value in excess of £400 million per annum and using live data from Transport
for London in apps can save users time to the economic value of between £15 million and £58 million
per annum.

While an aggregate figure on the social value of public sector information is difficult to reach, on the
basis of conservative assumptions, the figure could be in excess of £5 billion in 2011/12 (2011
prices). This figure is likely to increase as public sector information is used more widely and in more
impactful ways.

This yields a total economic and social value of the market for public sector information of between
£6.2 billion and £7.2 billion in 2011/12.

While a complete dataset on the number of public sector information datasets is currently not
collected, a preliminary estimated (based on a sample of data portals) suggests the volume of public
sector information datasets exceeds 37,500 – but this is likely to be an underestimate.

As discussed above, the narrow economic value of the market can be estimated to be between £1.2
billion and £2.2 billion in 2011/12.

The total economic and social value of the market can be estimated to be between £6.2 billion and
£7.2 billion in 2011/12.

There is currently insufficient reliable data to be able to make estimates on the size of the potential
market for public sector information in the future. However, the increased availability of better and
more public sector information, more linked data and better tools and techniques to exploit it mean the
value of the market is likely to grow.

The key gaps in the evidence base which, if filled, could lead to more precise estimates on the size,
include: (i) the total number of public sector information datasets collected across the public sector;
(ii) the individual cost elements behind each dataset and (iii) how public sector information datasets
are being used and re-used by customers and others downstream.

In the short term, these gaps could be addressed by (i) a more detailed survey of public bodies to
understand what datasets are being held and disseminated; (ii) a survey and consultation of
38
Market Assessment of Public Sector Information
Theme
Findings
customers of pay-for public sector information; (iii) a wider survey of users of open government data
or some form of limited tracking of how this information is re-used.


Can we assess the current and future for the; UK
market, EU market, Global market?
What indicators should we use to measure our
success in widening access to public sector
information?

One route to increasing the amount of data on the use and re-use of public sector information would
be to foster a climate of greater openness, taking into account commercial sensitivities around how
personal sector information is used/re-used. For example, one way of collaborating and fostering
more openness in future, providing it can be carried out in a non-too-onerous manner, might be
providing data free to third parties on the pre-condition that the information flow becomes a ‘two-way
street’ for Government and policymakers, e.g. “you can have our data, but we’d like to know how it is
being used and re-used to give us insight, benefit us and in turn UK society”.

This report has assessed the current market for public sector information in the UK and considered
how it might evolve. Many of these insights will be transferable to other markets and, subject to
overcoming any data limitations, it would be possible to assess the current and future potential of
public sector information markets overseas.

However, we note the potential issues arising from a straight read-across of outcomes from one
jurisdiction to another. As well publicised examples, public sector information value creation in
Denmark and New Zealand are widely cited as a reason for releasing more data. The UK is a larger
and more complex economy and data landscape, and as such the impacts of similar policy-changes
might be quite different.

Some organisations are already developing Key Performance Indicators around the quality and
format of data being made available.

However, success is ultimately determined by the ability of public sector information customers to be
able to effectively and efficiently access and exploit datasets as much as would be reasonably
expected.

Metrics could be created to measure progress against the barriers identified in this report.
Challenges to fully exploiting the value of public sector information, including issues around competitiveness, funding and regulation


What are the strengths and weaknesses of the UK
market?
What are the key issues affecting the data market?

The strengths of the UK market relate to the commitment by government to release more and more
public sector information and an emerging eco-system of companies and individuals able to exploit it.

The report has identified a series of barriers around legislation, economics and access which may
hinder the growth of the UK market.

The report has identified an issue around consumers’ and suppliers’ perceptions on the availability of
certain public sector information datasets. The report has also noted on-going complexities around
charging for certain datasets – the paucity of data on the total stock of public sector information being
held/collected, the different cost components behind individual datasets and how datasets are being
used and re-used by customers and the wider public, makes it difficult to make decisions around the
39
Market Assessment of Public Sector Information
Theme
Findings
provision of public sector information.

What are the medium and longer term trends in the
data market?

This lack of data (see above) is preventing accurate cost-benefit analyses and identifying where
scarce resources should be allocated to maximise the benefit from public sector information.

It seems reasonable to hypothesise that the larger part of the value to be generated from public sector
information lies in the future. Much of this value may come from combining public sector information
with information from other sources (e.g. held by the private sector), such as details of consumer
spending habits held by supermarkets and other retail firms, or details of domestic and commercial
properties.

Further benefits are likely to emerge for private individuals, business and other organisations as
methods of exploiting public sector information are developed and improved. More effective ways of
data exploitation can include:
o
greater system-wide analysis through more sharing and combining datasets together to
consider policy and business issues from a range of non-traditional perspectives;
o
unlocking the potential from linked data: and better integrating; and
o
greater data-mashing with private sector and individual datasets.

In addition, there are likely to be benefits from wholly new uses of public sector information.
Numerous additional possibilities for achieving such savings and improving policy may become
apparent as more information is made available and expertise in its use increases. The main report
summarises some examples of ‘blue skies’ thinking around public sector information.

How far does regulatory regime enable the market
and how far does it constrain it and in what ways?

Existing regulations and legislation governing public sector information do not appear to be acting as
actual barriers to realising the full potential value of public sector information. However, the manner in
which some of these regulations and legislation are interpreted can lead to overly risk averse
behaviour that can create barriers.

What licensing arrangements are currently in place
for PSI and how do they enable open data?

The Open Government Licence is widely regarded by stakeholders as an effective means of removing
barriers to exploiting the value of public sector information. However, this is not always used and does
not cover all public sector bodies.

How does the cost of public sector information affect
the data market?

At the broadest level, charging reduces the use of public sector information, but it can also act as a
signal of quality for (prospective) users.

The issue of charging for public sector information datasets and their funding models is complex.
While significant progress has been made by a number of public sector bodies in this area in
simplifying charging structures and making more public sector information available as open data,
there remains a perception that barriers exist, with the cost of datasets deterring use.

There is a lack of data (see above) to definitively conduct a cost-benefit analysis across different
funding models. This is an area for further analysis via detailed analysis of customer bases, costs and
40
Market Assessment of Public Sector Information
Theme
Findings
how data is used.

How clear are the standards that apply to public
sector information and how do they affect the
market?

Improvements continue to be made on public sector information datasets’ format and consistency,
with initiatives such as the Open Standards Principles and APPSI definition likely to contribute to this
– leading to common understanding and expectations around public sector information.

How can we categorise the innovative potential of the
market and how can we unlock that potential?

Given the current trajectory of increasing volumes of public sector information being made available in
ever more accessible formats, it seems reasonable to hypothesise that the larger part of the value to
be generated from public sector information lies in the future. Much of this value may come from
combining public sector information with information from other sources, such as details of consumer
spending habits held by supermarkets and other retail firms, or details of domestic and commercial
properties.

Many of these future uses may be extensions of the ways in which public sector information is already
being used. For example, a recent study calculated that the benefit to the local government sector
from the use of geospatial data was £232m over five years, through a variety of measures such as
more efficient routes for waste collection. It is to be expected that such uses will become increasingly
widespread throughout the public sector at both a local and central level, as more information
becomes available, methods of harnessing this information are improved, and best practice becomes
more widely adopted, thus increasing the value derived from public sector information.

Further benefits are likely to emerge for private individuals, business and other organisations as
methods of exploiting public sector information are developed and improved.

In addition, there are likely to be benefits from wholly new uses of public sector information.
Numerous additional possibilities for achieving such savings and improve policy may become
apparent as more information is made available and expertise in its use increases. In addition, there
are likely to be more ‘blue sky’ opportunities, such as those identified by the Government Office for
Science. These are explored further in Chapter 5, but include improving management and resilience
of food supplies, tackling obesity, and detecting and identifying infectious diseases. These require
combining public sector information with other sources of information rapidly and making the resulting
insights available to decision makers in a timely fashion to allow prompt responses to crises and
informed policy decisions.

Addressing the barriers identified in this study will certainly work to unlock this potential, but it is hard
to foresee specifically where innovation might take place in the UK. Often innovation takes place in
areas which are hard to predict, with first-movers benefitting accordingly. Quite possibly, the next data
innovators are already en-route to innovation on the back of public sectot information.
41
Market Assessment of Public Sector Information
1. Introduction and approach
This introductory chapter sets out the scope of this report, sets out its
approach and provides an outline of its contents.
Project scope and objectives
The Department for Business, Innovation and Skills on behalf of Stephan Shakespeare has
commissioned Deloitte 51 to undertake an independent assessment of the market for public sector
information, in order to establish a robust evidence base on its value and to highlight the policy
implications flowing from an examination of how PSI could be utilised further. The Deloitte report
forms the evidence base for the Shakespeare Review on public sector information.
When published, the Shakespeare Review will consider the full breadth of the public sector
information market, both current and future. It will consider how the private sector, civil society and
general public uses and re-uses public sector information, as well as the potential benefits for how
the public sector can use and re-use its own data. 52
The key elements of the Deloitte market assessment include:

a rapid evidence review to inform the definition of public sector information, users and reusers and what the market looks like and could look like in the future;

a desk-based data collection exercise setting out what the current and latent market for
public sector information looks like;

the construction of an analytical framework tracing based on the literature charting the routes
of impact publication sector information can have;

an ‘as is’ market analysis of public sector information in the UK, taking into account the
suppliers and competitors, consumers, nature of competition, value chain, regulatory
landscape, uses and re-uses of public sector information and identifying instances of market
failure;

three case studies examining how public sector information is used and re-used in the private
and public sectors. These include the health sector, use and re-use of public sector information
within government, and a ‘deep dive’ into the use of public sector information in the transport
sector;

high level quantitative analyses, where the data exist, assigning monetary estimates to the
current and potential value of public sector information; and

identifying policy implications on the different challenges to maximising the full potential of
public sector information.
51
Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited (“DTTL”), a UK private company limited by
guarantee, and its network of member firms, each of which is a legally separate and independent entity. Please see
www.deloitte.co.uk/about for a detailed description of the legal structure of DTTL and its member firms.
Deloitte MCS Limited is a subsidiary of Deloitte LLP, the United Kingdom member firm of DTTL.
This publication has been written in general terms and therefore cannot be relied on to cover specific situations;
application of the principles set out will depend upon the particular circumstances involved and Deloitte recommends that
you obtain professional advice before acting or refraining from acting on any of the contents of this publication. Deloitte
MCS Limited would be pleased to advise readers on how to apply the principles set out in this publication to their specific
circumstances. Deloitte MCS Limited accepts no duty of care or liability for any loss occasioned to any person acting or
refraining from action as a result of any material in this publication.
52
BIS, Independent Public Sector Information Review: Terms of Reference
Market Assessment of Public Sector Information
This report takes as its starting point that the public sector is equivalent to Category O of the
Standard Industrial Classification (SIC 2007) nomenclature – Public Administration and Defence;
Compulsory Social Security. However, when appropriate, public bodies outside this category (e.g.
health, transport and education) are considered and included.
The report covers all levels of government: UK, devolved, regional and local. Where appropriate, it
also considers the international dimension.
Approach
This report has adopted Deloitte’s standard five-stage methodological approach to develop the
evidence base.
Figure 1.1: Deloitte methodology summary
Source: Deloitte
This final report concludes the research.
Data sources
There is a paucity of data around public sector information in the UK making it difficult to generate
accurate figures on the size, value and potential of public sector information. This notwithstanding,
data on public sector information and other proxies has been collected from the following sources:

web-based PSI portals such as data.gov.uk and local government data stores and other
statistical agencies such as National Statistics;

existing literature on public sector information including the Open Data White Paper, Deloitte
research, studies by academics and other public bodies such as the Office of Fair Trading
(OFT);
43
Market Assessment of Public Sector Information

conversations with public sector information stakeholders such as the Open Data User
Group, the Trading Funds, regulatory bodies, government departments, business users and reusers and consumers;

annual reports of public sector information suppliers such as Trading Funds and departmental
open data strategies;

internal Deloitte research that draws on client experience and the use and re-use of public
sector information;

international comparisons such as from the EU, Australia and the USA.
As set out in Figure 1.1 this data was primarily collected between November and December
2012 53 . The approach to data collection has been to focus on gaining an overview of the public
sector information market at the broadest level, covering the widest possible range of issues and
underlying trends, noting that primary research was excluded from the project’s terms of reference.
Accordingly, all model results and data analysis should be read in conjunction with the relevant
caveats.
Structure of this report
This evidence base contains all of Deloitte’s analysis for this research project and is structured as
follows:

Chapter 1 – Introduction and approach: outlines the scope of this project and Deloitte’s
broad approach to the research;

Chapter 2 – Public sector information definitions and market definition: considers the
different definitions used for public sector information and goes on to discuss how public sector
information varies across a number of dimensions;

Chapter 3 – The supply and demand sides of the public sector information market:
provides an overview of PSIHs and some statistics on the supply of datasets, before moving to
examine how the datasets are used and re-used by businesses, individuals, community groups
and the public sector;

Chapter 4 – The regulatory landscape for public sector information: contains an overview
of the different regulations, guidelines and pieces of legislation that govern the operation of the
public sector information market;

Chapter 5 – The current value of the public sector information market: presents the
report’s calculations of the economic value of the market currently, as well as a series of case
studies that highlight the wider social value of public sector information;

Chapter 6 – Barriers to maximising the value of public sector information and their
policy implications: discusses the identified challenges to maximising the value of public
sector information in the UK; and

Chapter 7 – Conclusion: contains some closing thoughts on the subject.
The report’s appendices provide further methodological and analytical background. These are:

Appendix 1 – glossary: setting out the terms used in this research

Appendix 2 – acronyms used in this research

Appendix 3 – literature review: considers previous research in this area;
53
Although data continued to be received by the project team until March 2013.
44
Market Assessment of Public Sector Information

Appendix 4 – further statistics: in particular on the Trading Funds that make up the Public
Data Group;

Appendix 5 – empirical methodology: provides a full description of the assumptions and
approaches underpinning quantitative estimates;

Appendix 6 – transport detailed case study: contains additional details on the transport
centre case study methodology;

Appendix 7 – other case studies: additional details on various case studies conducted for
this research;

Appendix 8 – results from an informal Government Officials survey; and

Appendix 9 – bibliography.
45
Market Assessment of Public Sector Information
2. Public sector information
definitions and market definition
A key foundation stage in any market analysis is to establish which
products and services constitute the market under discussion. An
effective definition of public sector information and the parameters of
the market provide a strong framework for further analysis.
Definition of public sector information
Since their establishment, public bodies across the world have generated and retained a wealth of
data and information. The scale and range of this information can be overwhelming, as the
following examples demonstrate 54 :

HM Revenue and Customs interacts with over 40 million customers;

UK departments, agencies and other public bodies procure over £243 billion worth of goods
and services;

around six million people work in the public sector, each of whom has a record on
performance, sickness absences, skills and years of service; and

each year nearly 700,000 students make 2.7 million applications to university.
This data and information is typically collectively referred to as public sector information. However,
definitions of public sector information can, and do, vary. A review of the literature suggests that
there are a number of definitions in current use globally.
Figure 2.1: selected definitions of public sector information
Definition
Source
Information and data collected, reproduced and
disseminated by the public sector covering many
areas of activity, such as social, economic,
geographical, weather, tourist, business, patent and
educational information.
Re-use of Public Sector Information Regulations 2005
(SI 2005/1515), Reg.4(2) 55
Information, including information products and
services, generated, created, collected, processed,
preserved, maintained, disseminated, or funded by or
for the Government or public institution, taking into
account the relevant legal requirements and
restrictions.
OECD, Recommendation of the Council for Enhanced
Access and more Effective Use of Public Sector
Information [C(2008)36] 56
Information, data or content collected by and/or held
by a public body. The information may or may not be
OFT, The Commercial Use of Public Information
(CUPI), [OFT861] 57
54
Taken from Deloitte Analytics – Insight on tap: improving public services through analytics. Available at:
www.deloitte.com/view/en_GB/uk/industries/government-publicsector/67360f23824e0310VgnVCM2000001b56f00aRCRD.htm
55
Available at www.legislation.gov.uk/uksi/2005/1515/made
56
Available at www.oecd.org/internet/interneteconomy/40826024.pdf
57
Available at www.oft.gov.uk/OFTwork/publications/publication-categories/reports/consumer-protection/oft861
Market Assessment of Public Sector Information
Definition
Source
Crown copyright information.
Material with the essential purpose of providing
Government information to the public. Material with
the essential purpose of artistic expression is unlikely
to be treated as PSI for the purpose of the policy.
Office of the Australian Information Commission,
Principles on open public sector information: Report on
review and development of principles, (May 2011) 58
Information held by a public sector organisation, for
example a government department or, more generally,
any entity which is majority owned and/or controlled by
government.
Pollock, R, The Economics of Public Sector Information
(2008) 59
Note: in some cases the original definition has been summarised for purposes of brevity
As Figure 2.1 demonstrates, definitions of public sector information fall across a continuum ranging
from all information and data generated, held and disseminated by the public sector (content
agnostic) to the definition of public sector information explicitly covering certain types of data and
information (content specific). Importantly, all definitions are neutral with respect to the format of
public sector, whether it is qualitative or quantitative, structured or unstructured and whether or not
it is subject to user fees 60 .
For the purposes of this report, public sector information is defined as covering the wide range of
information that public sector bodies collect, produce, reproduce and disseminate in many areas of
activity while accomplishing their public tasks. This is consistent with the recently published
glossary of terms by the Advisory Panel on Public Sector Information (APPSI) 61 .
This report acknowledges that this definition is not entirely uncontested. In particular, there is
debate over when particular public sector information datasets become value-added services
rather than information and data. For example, Deloitte has heard views that the above definition of
public sector information is too narrow and fails to capture intangible advice given out to customers
that draws upon public sector information, but of which no record is kept 62 . However, these
reservations notwithstanding, the above remains a reasonable working definition for the purposes
of this evidence base and allows some parameters to be set around the analysis.
It is important to note that for the purposes of this report, as is conventional practice, the term
information (output of such processes that summarise, interpret or otherwise represent data to
convey meaning) is taken to include data unless otherwise specified.
Definition of the public sector
A corollary of defining public sector information is the need to define what is meant by the public
sector itself. At the most simplistic level, the public sector can be defined in reference to the
Standard Industrial Classification (SIC) codes; where the public sector is Category O of SIC(2007)
covering ‘public administration and defence; compulsory social security’ 63 . While this is a useful
starting point it does not capture other activities that might reasonably be considered as part of the
public sector such as education, health and transport services as well the public funding of other
activities such as the arts and other cultural bodies.
58
Available at:
www.oaic.gov.au/publications/reports/Principles_open_public_sector_info_report_may2011.html#_Toc293927686
59
Available at:www.rufuspollock.org/economics/papers/economics_of_psi.pdf
60
The definitions also capture case data collected as required by legislation, e.g. highways data
61
Available at: www.nationalarchives.gov.uk/appsi/appsi-glossary-a-z.htm#apps-p
62
For example, interpretation of datasets given over the phone.
63
See www.ons.gov.uk/ons/guide-method/classifications/current-standard-classifications/standard-industrialclassification/index.html for more details on SIC codes
47
Market Assessment of Public Sector Information
There is some debate as to whether data and information produced by cultural centres, such as
museums and libraries, should be considered public sector information, especially in cases where
these receive public subsidy. For the purposes of this report cultural institutions are considered to
lie within the public sector, unless otherwise specified 64 . For similar reasons, while not part of the
public sector, private and third sector contractors providing public services may be considered as
producers of public sector information. Universities and schools although not within the scope of
the PSI Directive are significant in public sector terms and are included in the definition of the
public sector for the purposes of this report.
Thus, for the purposes of this report, the public sector is taken to refer to include (but not
necessarily be limited to):

national and devolved government (Ministerial and non-Ministerial) departments and their
executive agencies;

Non-Departmental Public Bodies;

the National Health Service;

the Judicial Service;

the Armed Forces and Police Service;

Public Corporations and Trading Funds;

Independent Panels and Inquiries; and

Local Authorities.
In reality it is difficult to clearly define the boundaries of the public sector and the above list is not
exhaustive, but is deliberately broad to capture the wide range of public sector actors, ranging from
the departments of state, to the BBC and the NHS, through to local borough councils.
The nature of public sector information
As defined above, the term public sector information captures a vast range of information and data
gathered from diverse sources, for a wide range of purposes, and stored in many different formats,
with the single unifying feature that it is collected by public sector bodies. It includes data gathered
intentionally, the collection of which may be one of the organisation’s main tasks; and information
gathered incidentally while performing other functions. Public sector information thus has a number
of dimensions, as set out below
Figure 2.2: public sector information dimensions
Public sector information can be categorised along a number of non-competing dimensions:

whether the public sector information is made available to the general public and, if so, under what
conditions (if any);

the complexity of the dataset in terms of the number of records and variables, whether it is anonymised or
aggregated or is quantitative or qualitative (its verbosity);

how often the dataset is updated or replaced (its velocity);

whether the public sector information is generated or collected as part of an public body’s public task or
whether its generation is the result of other activities not related to public sector information;

the content of the public sector information dataset;

the electronic or non-electronic format of the dataset;

the ways in which the public sector information dataset is distributed; and
64
The revised EC PSI Directive (see chapter 4) also adds the cultural sector (archives, libraries and museums) within
scope.
48
Market Assessment of Public Sector Information

the cost of generating/collecting/maintaining/updating the public sector information dataset.
Source: Deloitte analysis
The following section defines these dimensions and explores each in more detail.
The availability of public sector information
It does not automatically follow that public sector information will always be made available to the
public or outside the public sector body that originally collected/generated it. Equally, it may not be
made immediately following its collection. For example, some data releases are not made public
immediately to avoid risking public safety, e.g. some Met Office data, which must be validated to
ensure it is correct; this is contrast to certain economic statistics which can be revised over time.
Figure 2.3 below summarises how public sector information can be retained and disseminated.
Figure 2.3: public sector information in the UK
Source: Deloitte analysis
It is important to highlight that public sector information released as open data is a sub-set of the
total public sector information market. To recall, public sector open data is data that has been
made available free of charge for anyone to use as they wish and meets the criteria of being
accessible, in a digital, machine readable format and free of restrictions on use or redistribution.
The ‘open definition’ is “a piece of data is open if anyone is free to use, reuse, and redistribute it —
subject only, at most, to the requirement to attribute and/or share-alike” 65 . The data portal
data.gov.uk contains a wealth of public sector open data 66 covering a range of topics ranging from
data on tariffs 67 to homelessness statistics 68 . Examples of data only available to the general public
under terms and conditions or for a fee include the National Pupil Database and subscriptions to
specialist statistics by the Office of National Statistics 69 . Data currently collected by the public
sector, but not made widely available outside the public sector, include many national security
information datasets.
The Open Government Licence 70 sets out the conditions under which individuals and users can
use and re-use PSI – though not all data and information available under the OGL this meets the
open data criteria (i.e. it may not be in a digital machine readable format). This report discusses the
regulations around public sector information in more detail in Chapter 4.
65
See http://opendefinition.org/
Much public sector information is now classified as open data, but the term open data itself is much broader, covering
any data – for example, data provided by private companies, individuals, and not for profit organisations – which is made
available free of charge to the public.
67
See http://data.gov.uk/dataset/uk-tariff-codes-2009-2010
68
See http://data.gov.uk/dataset/statutory_homelessness_statistics_england
69
See www.nomisweb.co.uk/
70
See www.nationalarchives.gov.uk/doc/open-government-licence/
66
49
Market Assessment of Public Sector Information
The complexity of public sector information
The level of detail of a public sector information dataset is referred to as its verbosity, with more
complex, larger datasets being more verbose. Verbosity may be a function of the number of
variables contained in the dataset, the number of records it holds and the time period it covers.
Public sector information datasets can either focus on aggregated data or be at an individual unit
level – the unit can vary between individuals, social groups, businesses, industries, economic
sectors and so forth – the more disaggregated, the more verbose.
The size of public sector information datasets can also vary considerably, with only a small
proportion being truly classed as Big Data, although this may not always be reported in the
statistics. For example, the Companies House dataset is substantially larger than, say statistics on
football banning orders, but both are counted as a single dataset by the official statistics. Equally,
the Met Office supplies around 200Mb data per day to data.gov.uk, but this is contained in just
three large datasets. The size of individual datasets also has implications as to how it is distributed
(see below).
The speed at which public sector information is updated
It is important to distinguish between public sector information that exists in a relatively static form,
and public sector information which is continually updated. The frequency with which a public
sector information dataset is updated is known as its velocity.
In some cases, public sector information datasets will only be updated infrequently, perhaps
annually or, in the case of the Census, every ten years. Typically, updates are made when an
additional cohort of data is available and the original dataset is extended to include this new data.
Examples of such datasets include information on departmental expenditure, educational
attainment of GCSE students, or numbers of welfare claimants – all of which are updated with
supplemental data at certain intervals (the time series expands). Other public sector information
datasets may also be updated periodically, but these updates replace existing datasets (there is no
time series of data) – such as the case of economic forecasts where the latest forecast supersedes
the previous one.
Other types of public sector information, however, are updated on a daily, hourly or on an even
more frequent basis and the usefulness of the public sector information is directly related to how
up-to-date it is. For example, Transport for London provides live travel information meaning that
updates on disruptions and delays, as well as bus and tube departure times, are updated in real
time. In these instances the need for the data to be regularly refreshed becomes of the utmost
importance, as historic data may have relatively little value. The updated dataset further needs to
be disseminated to users rapidly and in an appropriate format. This is commonly achieved through
an application programming interface, or API, which allows updates to be embedded in a webpage,
programme or mobile application. This information can be characterised as having a high velocity.
The distinction between the velocity of different public sector information datasets is significant
because the speed at which datasets are updated can impact their end use, methods of
dissemination and the level of commitment required from the PSIH. High velocity public sector
information is often collected by an organisation dedicated to that purpose, as in the case of the
Met Office and weather information. Given the time-sensitive nature of high velocity public sector
information, users are likely to demand a higher quality of service and assurance that the public
sector information will be up-to-date.
Public sector information as a by-product or purpose-collected
In some cases public sector information is generated in the course of a public sector body’s other
activities, rather than the information being generated or collected as its core activity. This is known
as ‘by-product’ or ‘exhaust’ data or information. This applies, for example, to the data contained in
the National Pupil Database; Home Office statistics on crime; salary information for employees of
public sector bodies; and details of public expenditure. The Department for Education collects data
on students in the course of its regular activities but the collection of this data is not its primary
50
Market Assessment of Public Sector Information
purpose – it arises from, and facilitates, other activities. The same may be said for salary
information published by government departments: the data is generated in the course of
employing people to undertake the range of functions required of the department.
The fact that this public sector information is a by-product of an organisation’s public task does not
necessarily imply that its value to users is in any way reduced. Nor does it mean that its collection
and dissemination is cost-free. There may be costs incurred in processing, storing and
disseminating the public sector information 71 . However, the cost of collection is generally thought
to be lower than purpose-collected data, since the public sector information is generated by
activities which would have taken place regardless of whether or not the public sector information
was collected 72 .
In other cases, the collection of public sector information is the purpose of a public sector
organisation’s activities and even its existence (part of its public service remit). This is the case, for
example, with Ordnance Survey. In these cases the public sector information has been identified
as desirable in its own right. It is also likely that this public sector information will be highly valued
by users, since it has been actively identified as worth the cost of collection.
The distinction between public sector information as a by-product of other activities and public
sector information that is purpose-collected is therefore significant, since it suggests differing costs
of collection and perhaps also different user profiles and, potentially, value.
The content of public sector information
Public sector information can also be categorised on the basis of content. These content
categories broadly follow the activities of the PSIH in question and cover all aspects of the UK’s
economic, social and cultural activities. In addition to specific datasets relating to the activities of
their staff and agencies, all PSIHs will also be generating public sector information in the form of
operational data and information covering the administrative and logistical activities of public sector
actors such as budgets, organisational charts, HR statistics and pay scales.
Some examples of public sector information content categories are shown in Figure 2.4.
71
In some cases, this may be significant if the data needs to be cleansed and re-structured to allow consumers to
effectively use and re-use it.
72
Costs will also vary according to how often the data is collected.
51
Market Assessment of Public Sector Information
Figure 2.4: selected public sector information content categories
Source: Deloitte analysis
Where public sector information datasets contain personal information they may be anonymised or
non-anonymised – if the latter, they are subject to restrictions on their release.
The format of public sector information
Public sector information can be downloaded in a wide variety of electronic file formats. At the time
of writing, there were nearly 50 different formats of electronic files available for download at
data.gov.uk (a reasonable proxy for public sector information currently available to the general
public, although certainly not a comprehensive source) – a number of which are proprietary
formats such as .pdf and .xls (even if the software readers are readily available).
The ten most popular file formats (constituting around 96 per cent of all datasets) were:
Figure 2.5: most popular public sector information file types on data.gov.uk
File type
Number of datasets (27th February 2013)
Percentage of total (27th February 2013)
.csv
2423
40%
.xls
1624
27%
.pdf
588
10%
.html
461
8%
.rdf
191
3%
.xml
189
3%
.zip*
123
2%
.wms
69
1%
.doc
59
1%
.txt
52
1%
Source: Deloitte analysis
*Once unzipped, the file could be any of the above other files.
52
Market Assessment of Public Sector Information
Figure 2.4 refers to the format of public sector information currently available online at the time of
writing on data.gov.uk 73 . However, it is likely that there is public sector information currently
retained (or even collected) by the public sector which has yet to be digitised and is not available
electronically (such as archive and historical material 74 ). This public sector information may or may
not be accessible to the general public, but can only be provided in hard copy or cannot be
removed from site.
In addition, much public sector information is released in the form of an API 75 . This allows
developers to embed a data source in a webpage, programme or mobile application, allowing for
updates to the data. Prominent examples of public sector information APIs include Transport for
London, which provides live travel updates delivered via a webpage or mobile applications, and the
Met Office weather information, which is periodically updated with the latest data.
Star rating of public sector information
Related to the issue of format is the usability of the dataset. There are different ways in which
usability can be measured, and they are, to some extent, subjective. One of the more common
methods is the Open Data five-star scoring mechanism suggested by Sir Tim Berners-Lee, which
is used by data.gov.uk.
Figure 2.6: open data scoring
Score
Description
Make your stuff available of the Web (whatever format) under an open
licence
Make it available as structured data (e.g. Excel instead of image scan of
a table)
Use non-proprietary formats (e.g. CSV instead of Excel)
Use URLs to identify things, so that people can point at your stuff
Link your data to other data to provide content
Source: http://5stardata.info/
At the time of writing data.gov.uk has recently begun to provide data on the proportion of datasets
attaining each star rating (currently in beta form). These figures indicate that, at the time of writing,
nearly a quarter of datasets have attained a three-star rating. Only one per cent of datasets have
been given a five-star rating, indicating they are linked data. Over 50 per cent of datasets have
attained no stars – though this is due to many datasets not yet being rated 76 or the link referring to
data such as live mapping services (for which a rating has not yet been agreed as they are not
strictly data).
73
It does not include other data portals such as legislation.gov.uk.
Though some archive material, such as that held by The National Archives, can digitised on demand and electronic
copies supplied if a user wishes to pay for the service.
75
A specification intended to be used as an interface by software components to communicate with each other. An API
may include specifications for routines, data structures, object classes, and variables.
76
This may be because the algorithm to award the star rating to datasets is unable to categorise certain datasets
effectively.
74
53
Market Assessment of Public Sector Information
Figure 2.7: star ratings of datasets on data.gov.uk
Score
Description
Number of datasets (27th
February 2013)
Proportion
TBC
N/A
842
9%
No stars
Unavailable
5117
56%
Unstructured data (e.g. PDF)
163
2%
Structured data but proprietary
format (e.g. Excel)
826
9%
Structured data in open format
(e.g. CSV)
2004
22%
N/A
N/A 77
N/A
Linked data - data URIs and linked
to other data (e.g. RDF)
107
1%
Source: wwww.data.gov.uk/data/search
The distribution channels of public sector information
The analysis undertaken for this report suggests that the key distribution route of public sector
information by suppliers (or originators) is through websites 78 . The next chapter considers in more
detail the different websites from which public sector information is assessed, but broadly speaking
the ways in which public sector information is accessed and disseminated in the first instance are:

through viewing the public sector information online on a webpage;

through downloading the public sector information from a webpage;

through an API embedded in an app or webpage;

via secure terminal facilities or virtual laboratories to view the data 79 ;

other published versions of the work; and

through requesting the public sector information via a webpage, telephone or other written
request. The public sector information may be sent via email, an encrypted link, CD or DVD or
hard copy.
Often the latter case is imposed for public sector information datasets that are too large to
efficiently host online, that contain sensitive personal data or are subject to copyright conditions
and which can only be released upon the receipt signing licensing terms and conditions. For
example, given the nature of the data and information contained within it, the National Pupil
Database can only currently be shared with researchers following the signing of specific terms and
conditions. It should be noted that in any case this data can only be shared where there is a
statutory gateway to permit this (as is the case with the National Pupil Database).
77
Not included in the data
Though, as discussed above, there is likely to be a wealth of historic public sector information that is not readily
available online.
79
For example, the ONS allows researchers access to micro-data in secure facilities.
78
54
Market Assessment of Public Sector Information
The cost of public sector information
Public sector information differs in terms of cost. For certain types of public sector information the
marginal cost of generation/collection may be relatively high (compared to dissemination) if this
data is being collected specifically. Examples of the former include the geospatial data collected by
Ordnance Survey, which requires use of aircraft and personnel on the ground. On the other hand
the data collected by Transport for London arises from its day to day operations, and producing
this data requires relatively little investment of resources beyond the infrastructure already in place
to support operations, meaning that there is a low marginal cost of generation – although the costs
of dissemination may be higher.
The public sector information cost lifecycle components are shown below.
Figure 2.8: public sector information cost lifecycle
Source: Deloitte analysis. Costs of acquisition may also include purchasing third party data that is combined with public sector
information
Public sector information can be released to the general public at different parts of the cost
lifecycle – PSIHs do not necessarily have to wait until the public sector information has been
cleansed (data assured) and formatted into refined data before they more widely disseminate it.
The marginal cost at each point of the lifecycle will differ, e.g. the marginal cost of dissemination
may be very different to the marginal cost of collection. PSHIs can release public sector
information prior to cleaning as unrefined or raw data.
How public sector information is funded is discussed in more detail in subsequent chapters. Where
public sector information datasets do carry charges these can be in form of one-off fees, ongoing
subscription fees or a royalty model.
Having defined public sector information, the next section considers its market definition.
Market definition
Before considering the actors on the supply, intermediary and demand sides of the market, a
preliminary analytical step is to define what is meant by the market for public sector information. As
55
Market Assessment of Public Sector Information
set out by the OFT in its guidance for market definition 80 , the process of defining a market typically
begins by establishing the closest substitutes to the product or service in question. These
substitute products are the most immediate competitive constraints on the behaviour of the
organisations supplying the product or service in question 81 . As highlighted in an earlier chapter, in
the case of public sector information the market definition exercise is complicated by the fact that
different public sector information datasets vary substantially by content, quality, size and format.
As discussed in previously, at a first glance, the overwhelming majority of public sector information
is provided by the public sector through PSIHs.
Figure 2.9: examples of public sector information web portals
However, it is not enough to stop at the PSIHs when considering who the suppliers of public sector
information are – there may be substitute suppliers outside the public sector. Figure 2.10 highlights
a selection of public sector information datasets, their suppliers and potential substitutes.
Figure 2.10: public sector information datasets and substitutes 82
Public sector information dataset
Public sector supplier
Potential substitute supplier (if
any)
UK mapping data
Ordnance Survey
Google Maps, Open Street Map
Meteorological data such as weather
forecasts
Met Office
Various private sector suppliers or
overseas suppliers and academic
institutes
Economic statistics
National Office of Statistics,
Bank of England
None currently
Source: Deloitte analysis
80
See Market Definition: understanding competition law (OFT403) available at
www.oft.gov.uk/shared_oft/business_leaflets/ca98_guidelines/oft403.pdf
81
Although some of these substitutes may come with greater restrictions over the use and re-use of public sector
information.
82
The focus here is on substitutes for the underlying data in the public sector information dataset not the value-added
service that may be associated with it (including intermediary services).
56
Market Assessment of Public Sector Information
What this short comparison shows is that the presence of alternative suppliers varies with content,
with some types of content being more readily substitutable: this could include mapping data or
meteorological data. Of course, whether Google Maps is an effective substitute for Ordnance
Survey will depend on whether consumers regard them of substitutable quality. This can be tested
through the so-called hypothetical monopolist test 83 .
In this case, taking the Met Office as an illustrative example, the question would be whether the
Met Office could profitably raise prices for meteorological data by 5-10 per cent above competitive
levels. If the answer is yes, then the test is complete and the market for meteorological data solely
consists of the Met Office. If the answer is no, then this is due to consumers switching to other
substitute products such as rival suppliers of meteorological forecasts (assuming they exist). In this
case, the hypothetical monopolist test is repeated, assuming the Met Office also controls overseas
suppliers. If it can profitably raise prices by 5-10 per cent then the market is defined; if not, the
process is repeated with more substitute suppliers until the market is defined.
The ability of this report to carry out an empirical hypothetical monopolist test has been constrained
by a lack of elasticity data on potential public sector information substitutes. Further it is not
practical to carry out the test for the thousands of different public sector suppliers of public sector
information. Accordingly, as a working market definition, this report broadly defines the market for
public sector information as including only public sector suppliers, unless there are strong reasons
to include other non-public sector suppliers (this may be the case for certain types of geo-spatial
and mapping data).
The geographic market definition is restricted to the UK.
Chapter summary

For the purposes of this report, public sector information is defined as covering the wide range of information
that public sector bodies collect, produce, reproduce and disseminate in many areas of activity while
accomplishing their public tasks.

For the purposes of this report, the public sector is taken to refer to (but not be limited to):

o
national and devolved government (Ministerial and non-Ministerial) departments and their executive
agencies;
o
Non-Departmental Public Bodies;
o
the National Health Service;
o
the Judicial Service;
o
the Armed Forces and Police Service;
o
Public Corporations and Trading Funds;
o
Independent Panels and Inquiries;
o
Local Authorities.
Public sector information varies in terms of:
o whether it is made available to the general public and, if so, under what conditions (if any);
o its complexity in terms of the number of records and variables, whether it is anonymised or
aggregated or is quantitative or qualitative (its verbosity);
o how often it is updated or replaced (its velocity);
83
The OFT defines this as a “test that seeks to establish the smallest product group (and geographical area) such that a
hypothetical monopolist controlling that product group (in that area) could profitably sustain 'supra competitive' prices, i.e.
prices that are at least a small but significant amount above competitive levels. That product group (and area) is usually
the relevant market. If, for example, a hypothetical monopolist over a candidate product group could not profitably sustain
supra competitive prices, then the candidate product group would be too narrow to be a relevant market. If, on the other
hand, a hypothetical monopolist over a subset of a candidate product group could profitably sustain supra competitive
prices, then the relevant market would usually be narrower than the candidate product group.” Available at http://www.oft.gov.uk/shared_oft/business_leaflets/ca98_guidelines/oft403.pdf
57
Market Assessment of Public Sector Information
o
o
o
whether it is generated or collected as part of an public body’s public task or whether its generation
is the result of other activities not related to public sector information;
by its content;
its format;
the ways in which it is distributed; and
o
the cost of generating/collecting/maintaining/updating it.
o

This report broadly defines the market for public sector information as including only public sector suppliers
unless there are strong reasons to include other non-public sector suppliers. The scope of the analysis is
restricted to the UK.
58
Market Assessment of Public Sector Information
3. The supply and demand sides of
the public sector information
market
This chapter provides an overview of the supply and demand sides of
the public sector information market. It provides statistics around the
supply of datasets and profiles the different types of public sector
information holders (PSIHs).
The chapter then moves to consider intermediaries and the final
consumers on the demand side of the market. It looks at the public
sector information value chain, different business models in operation
and the ways in which public sector information is currently used and
re-used.
Supply side of the market
As highlighted in the preceding chapter, public sector information is collected and/or generated by
thousands of organisations across the public sector – so-called Public Sector Information Holders
(PSIHs). These PSIHs can be found in UK, devolved, regional and local government, as well as
other public sector bodies such as the armed forces, police force, NHS, universities and others
named in Chapter 2.
Supply of datasets nationally
When considering the identity of the main suppliers of public sector information it is important to
distinguish between PSIHs who hold the most information and PSIHs who make the most
information available to the general public. As an example, the Ministry of Defence and the
Security Services may potentially hold the largest amounts of public sector information of any
PSIH, but they may be one of the smallest suppliers of information to the market (either as open
data or under restrictions). Given the lack of robust data as to the overall size of the public sector
information market (i.e. public sector open data + public sector information available under
restrictions or for a fee + public sector information collected/generated not disseminated to the
general public), this report can only seek to make an approximation as to who the largest suppliers
of public sector information are in terms of information made available to the public.
With respect to open public sector information 84 , at a national level, much of this public sector
information is made available through data portals, specifically data.gov.uk (the single largest
source of public sector open datasets). An analysis of this provides a useful, illustrative tool for
assessing the most prolific suppliers of public sector information amongst PSIH. An initial
assessment shows that of a total of 781 publishers (as of 27th February 2013), the top ten suppliers
of PSI supplied over half of all datasets published on the website.
The top ten data.gov.uk publishers are as follows:
84
There is less data available on the number of datasets that are not available as open data or under the Open
Government Licence (see Chapter 4). However, particular instances of datasets available under licence or for a free are
discussed in this and subsequent chapters and appendices.
Market Assessment of Public Sector Information
Figure 3.1: PSIHs with the largest number of datasets published on data.gov.uk,
February 2013
Supplier
Number of open
datasets (as of
27/02/13)
Proportion of all
datasets on
data.gov.uk
Office for National Statistics
847 85
12%
Department for Communities and Local Government
744
11%
Health and Social Care Information Centre
589
9%
British Geological Survey
379
6%
Department for Environment, Food and Rural Affairs
331
5%
Centre for Ecology & Hydrology
307
5%
Welsh Government
241
4%
Department of Health
240
4%
Department for Children, Schools and Families
227
3%
Home Office
213
3%
Total
4,118
60%
Source: Deloitte analysis of data.gov.uk
The content of the datasets available on data.gov.uk is also summarised below.
Figure 3.2: data.gov.uk dataset links, October 2012
Source: Deloitte analysis of data.gov.uk
The numbers in Figures 3.1 and 3.2 should be read carefully as they refer to the number of
datasets only and by download links 86 . Furthermore, they do not take account of the velocity and
verbosity of the information.
As an example, as part of the evidence gathering process for this report, the Met Office has
indicated that their daily submission to data.gov.uk amounts to around 200Mb of data, which
undoubtedly makes them one of the largest suppliers of public sector information to data.gov.uk.
85
This figure appears to refer to the main datasets (e.g. Census and Labour Force Survey) – as clearly, the ONS
publishes significantly more individual pieces of data.
86
I.e. there may be more than one dataset per download link so the numbers are likely to be an underestimate.
60
Market Assessment of Public Sector Information
However this importance is disguised by the fact that the data/information is contained in just three
datasets 87 . It is therefore important to treat information on the number of datasets with a degree of
caution: it is a reliable guide to neither volume nor value of public sector information. Nonetheless,
as a first approximation, the analysis is useful to illustrate the long tail of PSIHs: whereas around
half of all PSI datasets come from 10 PSIHs, the remaining datasets come from over 750 PSIHs.
Moving away from data.gov.uk, open public sector information is also available from a number of
other websites. Deloitte’s recent research with the ODI has examined the number of open datasets
types (rather than links) available across three major data portals: data.gov.uk, the ONS and
data.london.gov.uk – this is shown below.
Figure 3.3: number of open datasets across data.gov.uk, the ONS and
data.london.gov.uk 88
Source: Deloitte / ODI analysis of various public sector information data portals
The number of links / datasets in both Figure 3.2 and 3.3 is dominated by government spending
information and data. In some senses this is unsurprising as this type of public sector information is
likely to have relatively low cost of collection and dissemination for PSIHs, and is relatively easy for
them to make available online on data.gov.uk, subject to any data redactions. 89 PSIHs have also
been incentivised to publish this type of public sector information under the Government’s
transparency agenda 90 .
Other sources of public sector information available (at the time of writing) as open data include 91 :

Companies House – with one major free open dataset;
87
Indeed, the Met Office’s daily files include modelling and observational data for over 5000 sites across the UK ingested
by data.gov.uk which are then converted, per location, into a database row with observational and modelling data. These
three files generate 250 million lines of data each hour.
88
Another example of a website counting the number of datasets is the TWC LOGD Linking Open Government Data
website: http://logd.tw.rpi.edu/iogds_data_analytics
89
This is because it is largely ‘by-product’ data: for example, information on salaries, spending by departments, and
organograms
90
See for example: www.cabinetoffice.gov.uk/news/government-spending-data-published
91
This list is not exhaustive.
61
Market Assessment of Public Sector Information

The Met Office – with 118 open data sets (listed on data.gov.uk);

Land Registry – with over 30 open data sets (available through data.gov.uk);

Data.london.gov.uk – with over 550 open data sets;

OS OpenData – with 12 mapping products that can be ordered online;

Open Data Communities – with over 70 open data sets; and

Legislation.gov.uk – with over 78,700 open data sets 92 .
A separate statistics appendix 93 provides further detail on datasets held by the four Trading Funds
that make up the Public Data Group.
Downloads and popularity of datasets
Using data.gov.uk statistics again, it is also possible to examine the popularity of different public
sector information open datasets in terms of page views.
Figure 3.4: data.gov.uk dataset normalised page views to show relativity, October
2012
Source: Deloitte analysis
The equivalent chart for dataset downloads from a broader group of websites is shown below.
92
Individual acts etc. This figure is approximate and may not include all PSI available of legislation.gov.uk. There is an
argument that this should be considered a single dataset, but given that the files are downloaded separately and not as
one database they are here treated as separate datasets.
93
See Appendix 4.
62
Market Assessment of Public Sector Information
Figure 3.5: open data downloads
Source: Deloitte / ODI analysis of various public sector information data portals
As one might expect, business and economy datasets dominate, although this may simply be a
function of the supply of these datasets exceeding those of other public sector information types.
However, data on social conditions and transport also report high levels of relative popularity.
Between July 2012 and January 2013 there were over a million total visits to data.gov.uk and over
3.7 million page views 94 . While this is a high level of traffic, many of these users would not have
downloaded data (see below). In some cases this may simply have been due the visitor deciding
they no longer wished to see the data or automated hits from web-trawlers. However, in other
cases, this may be because individuals were unable to find the particular dataset they were looking
for; or because they encountered broken links; or that they were transferred to the original PSIH’s
website to download the data rather than download it from data.gov.uk. Without a survey of users it
is difficult to know the reasons why a visit ended with no dataset download. However, the National
Audit Office found that 82 per cent of users left data.gov.uk directly from the homepage or data
page, indicating that they may have not found the information they were seeking. 95 Note, this was
a snapshot figure in early 2012 and data.gov.uk continues to evolve.
Figure 3.6 below indicates the most downloaded datasets from data.gov.uk between July 2012 and
January 2013. In total there were over 80,000 downloads of datasets from data.gov.uk during this
period. 96 While this is a significant number, it means that at most eight per cent of visitors
downloaded a dataset – and fewer if some users downloaded more than one dataset, as seems
likely.
94
Source: www.data.gov.uk/data/site-usage#totals
NAO, Implementing Transparency, April 2012. Available at www.nao.org.uk/wpcontent/uploads/2012/04/10121833.pdf
96
Source: www.data.gov.uk/data/site-usage/dataset
95
63
Market Assessment of Public Sector Information
Figure 3.6: data.gov.uk most downloaded datasets (July 2012 – January 2013) 97
Source: www.data.gov.uk/data/site-usage/dataset
Of course, these statistics do not take into account views and downloads from other PSIHs
websites and non-public sector PSIHs. Further, these statistics do not include public sector
information sent to data.gov.uk but hosted on other platforms. For example, the Met Office data
supplied to data.gov.uk is hosted by Microsoft Windows Azure Datamarket 98 - approximately 250
million rows of data per hour is hosted on this service and it is estimated that there are currently
over half a million transactions per month, on average. Nearly two thirds of these transactions
come from outside the UK suggesting UK public sector information can also help generate growth
overseas.
Taking these into account, the numbers of total downloads of public sector information may be
much higher.
Supply of datasets locally
As well as open public sector information being available at the national level, consumers can
download and access local open public sector information. The Local Government Association
carried out a survey of local authorities in late 2012 (‘Local Government Transparency Survey
2012’) 99 to examine the provision of data at a local level. 113 local authorities responded out of
97
"Downloads" is the number of times a user has clicked to download either an original or cached resource for a
particular dataset. Download information is only available from 2nd December 2012 (http://data.gov.uk/data/siteusage/dataset)
98
See https://datamarket.azure.com/dataset/0f2cba12-e5cf-4c6d-83c9-83114d44387a
99
See www.local.gov.uk/web/guest/local-transparency/-/journal_content/56/10171/3825698/ARTICLE-TEMPLATE
64
Market Assessment of Public Sector Information
128 respondents to the survey (the total surveyed was 346). The chart below shows the types of
information that respondents reported they currently publish or plan to publish in the future (per
cent).
Figure 3.7: Local Government Transparency Survey 2012 results – public sector
information published and plans to publish (%)
Source: LGA, ‘Local Government Transparency Survey 2012’
The data being made available by local government is often not published through a national data
portal such as data.gov.uk. The survey revealed that of 104 respondents to the question, 14 per
cent made information available through data.gov.uk and nine per cent through direct.gov. The
table below shows the full responses to the question ‘Where on the internet do you publish your
open data/transparency data?’
Figure 3.8: Local Government Transparency Survey 2012 results – public sector
information channels
Publication channel
Number
Per cent
responding
Dedicated open/transparency data page
71
65
Directly within topic section pages
53
48
Data.gov.uk
15
14
Direct gov.uk (now disbanded)
10
9
Publication schema
23
21
Information asset inventory
0
0
65
Market Assessment of Public Sector Information
Publication channel
Number
Per cent
responding
Other
10
9
Source: LGA, ‘Local Government Transparency Survey 2012’
Because local authority information is published through many different channels, it is difficult to
get a sense of the number and types of users of the information that has been released.
Nonetheless, the LGA survey offers some insight in this regard. There were 104 responses to the
question ‘Do others use your open data?’ The answers reveal that 16 per cent of local authorities
were certain that others used their open data, while nearly half believed they did not. 36 per cent
were unsure, indicating that it is often difficult to track how many users of open data there are,
especially where this does not require registration by users.
14 local authorities responded that they knew who used their open data. Their answers are shown
in the table below, showing that local community groups are particularly significant users of local
authority open data. However, given the small sample size, these results should be treated with
some caution.
Figure 3.9: Local Government Transparency Survey 2012 results – public sector
information consumers
Users of open data
Number
Local community groups
7
Other councils/public services
3
Individuals
3
Local educational establishments
2
Local charities
1
Local businesses
1
Source: LGA, ‘Local Government Transparency Survey 2012’
Trading Funds
Trading Funds are among the most visible PSIHs in the wider public sector information landscape,
and are often the focus of attention in debates over both the value of public sector information and
access for users 100 . While this report focuses on the wider public sector landscape, of which the
Trading Funds are only one component, given the nature of the public sector information that the
Trading Funds collect, as well as the debate over issues around access, it would be remiss not to
separately consider their role in the supply of public sector information.
While no Trading Fund is identical, collectively they occupy a special position in the public sector
information landscape as bodies with a statutory requirement to fund their operations and meet
their financial objectives from trading income. As such they are subject to a set of guidance
documents, legislation and regulations and guidance that differ from most other public sector
bodies. The key elements of this framework are described below (this list is not exhaustive and
these do not exclusively apply to Trading Funds).
100
These debates are discussed in depth in Chapters 5 and 6.
66
Market Assessment of Public Sector Information
Figure 3.10: regulations and guidance covering Trading Funds 101
Regulation / guidance
Content
Managing Public
Money (MPM)
The PDG Trading Funds are required to comply with the guidance on fees,
charges and levies in Managing Public Money. The standard approach to setting
charges for public services (including services supplied by one public sector
organisation to another) is full cost recovery. This normally means recovering a
3.5 per cent real charge for the cost of capital. However, for services supplied into
competitive markets, charges should be set at a commercial rate.
Managing Public Money explains that the norm is to charge full cost for publicly
provided goods and services. For commercial services, charges should be set at
a commercial rate. Managing Public Money also explains that much information
about public services should be made available either free or at low cost in the
public interest. However, there are circumstances where charges are made.
Public sector organisations can also charge for information which recipients
intend to re-use. Managing Public Money explains that where data is supplied for
re-use, the norm is to charge marginal cost. For value added data, and for all
information supplied by Trading Funds, the norm is to charge at full cost plus an
appropriate rate of return.
European Union
Directive on Re-use of
Public Sector
Information
Under the 2005 legislation, charges for the re-use of information made by public
sector bodies cannot exceed the cost of collection, production, reproduction and
dissemination of documents; and a reasonable return on investment.
The revision of this directive is proposing to move to marginal cost but allows for
some exceptions.
Information Fair
Trading Scheme (IFTS)
This sets and assesses standards for public sector bodies. It requires them to
encourage the re-use of information and reach a standard of fairness and
transparency. The National Archives use the IFTS to monitor the activities of the
trading funds to ensure that they trade fairly, openly and transparently in
information. The IFTS also covers other public bodies.
Source: Public Data Group, ‘Approach to Charging’ (December 2012)
For clarity, it should be noted that Trading Funds have their own Trading Fund Order – statutory
duties to charge are only in place for the Land Registry and Companies House, but the Trading
Fund Order and Trading Fund Act set out the framework for all Trading Funds’ operations.
It is important to be clear that these guidance documents and regulations, and especially the
requirement to recover costs plus 3.5 per cent, constrain the ability of the Trading Funds to change
their underlying business models beyond a certain point.
There are a number of Trading Funds currently operating in the UK, each arising out of distinctive
historical circumstances and each with a very different purpose, remit and business model. Below,
an overview of the four Trading Funds that constitute the Public Data Group (PDG) is provided.
This draws on the work of the Shareholder Executive, December 2012 102 .
101
Note: the below is adapted from a Public Data Group paper, ‘Approach to Charging’, December 2012. This is
available at www.gov.uk/government/publications/approach-to-charging. The regulations here discussed do not
necessarily apply exclusively to trading funds.
102
Ibid.
67
Market Assessment of Public Sector Information
Figure 3.11: Public Data Group Trading Funds
Companies
House
Land Registry
Met Office
Ordnance Survey
Status
Trading Fund
Statutory Body
Trading Fund
Statutory Body
Trading Fund
Trading Fund
Competitive
Landscape
100% noncompeted
(statutory
monopoly)
98% noncompeted
(statutory
monopoly)
Mix of:
All products and
services are traded in
competed markets
(apart from OS
OpenData).
 Non-competed and
competed HMG
contracts; and

Widely competed
commercial (nonHMG) contracts.
Sources of
revenue
Registration
and search
customers pay
statutory fees.
Data users
fund
dissemination
costs.
Register users
fund data
collection through
statutory fees.
Value added
service users
fund product
development and
dissemination.
HMG funds the core
underpinning national
infrastructure, research
and development
required for the national
weather, core climate
services and services to
defence.
Wholesale data users
fund their own data
dissemination costs.
Users of value added
competed services pay
for the bespoke
development of weather
services and the transfer
price of the data, which is
on the same terms for all
parties.
Data users fund data
collection, product
development and
dissemination.
Users of value-added
services pay for the
bespoke development
of services and the
transfer price of the
data, which is on the
same terms for all
parties.
Public and
private
sector users
99% non-HMG
1% HMG
100% non-HMG
79% Core HMG
4% Competed HMG
17% non-HMG
43% PSMA, OSMA
43% other Chargeable
incl. Private Sector
14% OpenData™ (for
private and public sector
innovation)
(1)
(1) PSMA: Public Sector Mapping Agreements; OSMA: One Scotland Mapping Agreement
Source: Shareholder Executive
As described above, Trading Funds are required to generate their own revenue, which is to say
that they do not in general receive centrally allocated funds. They are required to recover the full
cost of service provision plus the cost of capital, set at 3.5 per cent. The exception is where
Trading Funds operate in competitive markets, in which case they are required to price their
services at market rates to avoid distorting competition. Even where Trading Funds provide
information or services to public sector organisations, they ‘sell’ these at agreed rates. This is an
important distinction, making the Trading Fund business models qualitatively different to that of
other publicly funded PSIHs which provide ‘free’ services to public sector users (though certain
services may be charged for). The Trading Fund model is intended to drive efficiency by forcing
Trading Funds to hold down costs in the same way as a commercial business would do.
Further detailed statistics on the amount of data being made available and under which charging
models is provided in Appendix 4.
68
Market Assessment of Public Sector Information
Attitudes and policies of public sector information holders
As part of wide-ranging stakeholder consultations, Deloitte conducted an informal consultation of
Government officials across national departments to seek to develop indicative estimates of the
amounts of public sector in amounts of public sector information currently held and currently made
available and understand some of the challenges currently faced. Full details of the consultation
and its results can be found in Appendix 8 103 .
The most important findings from the survey included:

departments currently vary significantly in the amount of public sector information they make
available to the public;
o
around a third of respondents to the informal survey indicated that up to a quarter of
public sector information held by the department (and in some cases its executive
bodies) was currently being made available to the public. Other respondents were
unable to give estimates

of the public sector information made available to the public, all respondents said it was
available either entirely free of charge or at least 75 per cent was available at no cost; and

33 per cent of respondents report that less than five per cent of their staff were dedicated to
data collection, processing and dissemination (though a much higher proportion would be
involved as part of their wider duties), while 50 per cent were unable to estimate the proportion.
The exceptions were dedicated statistical bodies, which owing to their nature employ many
more staff dedicated to data-related functions.
All respondents indicated that their PSIH planned to make more public sector information available
to the public over the next 12 months, as set out in departmental open data strategies.
The officials consulted in this informal consultation work with public sector information issues on a
regular basis and are likely to have had a high degree of awareness of public sector information
issues. A recent survey by Listpoint 104 surveyed a wider number of civil and public servants to
examine levels of awareness of the benefits of open data and public sector information initiatives.
The key highlights are shown below:

78 per cent of those surveyed did not know about government plans for open data and the
benefits that follow;

57 per cent of those surveyed did not know how to access datasets, how to interpret them or
how to best apply data standards;

Over 75 per cent of those surveyed did not know their role in delivering the open data agenda;
and

72 per cent recognised that knowing how to access, share and use data will be increasingly
important over the next three years.
Again, it should be noted that there may an element of self-selection in the survey results.
103
This informal survey should not be treated as statistical robust as there is no general public sector asset register to
accurately capture the size of the market, but it does provide a useful indication of some of the challenges faced. While
responses were received from a number of departments, responses were not received from some major departments
which, a priori, are expected to hold significant amounts of public sector information.
104
See: www.intellectuk.org/media-centre/member-press-releases/8967-first-benchmark-survey-of-open-dataunderstanding. The research was conducted in late December 2012 and more data is available on Listpoint’s website:
www.listpoint.co.uk/media. More than 1000 responses were received across central and local government, nondepartmental bodies, the NHS and police forces.
69
Market Assessment of Public Sector Information
Demand side of the market
Consumers on the demand side of the public sector information market can be divided into two
broad categories based on whether they use or re-use public sector information. Use of public
sector information refers to exploiting the data and information in a way that is commensurate with
the original purpose within which the public sector information was produced. An example might be
using economic statistics produced by the ONS to inform economic policymaking and investment
decisions. In contrast, re-use of public sector information refers to the use of the dataset in a way
other than the initial purpose for which it was collected/generated. An example might be correlating
data on environmental and atmospheric conditions to the academic performance of school
children. Clearly, the distinction between use and re-use is not hard and fast, as some aspects of
use could be interpreted as re-use.
Ultimately, when considering final consumers, almost every inhabitant of the UK is, at some point,
a downstream user of public sector information in some form or other – often indirectly where the
data is an input into another value-added service or product. This might be through a transport
information or weather forecasting mobile app, an Ordnance Survey map, or another product that
relies on public sector information to function.
The different types of consumers of public sector information can be charted to form a circular
value-chain. As Figure 3.12 highlights, the role of intermediaries in making public sector
information accessible to a wider audience is currently crucial. One might expect that as general
awareness of public sector information increases and the capacity to perform analysis directly
increases, the role of intermediaries may diminish.
Figure 3.12: public sector information eco-system
Source: Deloitte analysis
70
Market Assessment of Public Sector Information
The supply of public sector information out of PSIHs is not a one-way street - consumers and
infomediaries can request more datasets to be released via the data.gov.uk request mechanism 105 .
As discussed above, there is a paucity of statistics on public sector information users and re-users.
The following sub-sections present the available statistics on use and re-use in terms of download
statistics, consumers perceptions of the market, different business models in operation and a
profile of the emerging data scientist role.
Public sector information use and re-use
Download and usage statistics
One approach to measuring the extent to which public sector information is re-used is to consider
the number of apps that take currently available public sector information as an input. Figure 3.13
below charts the number of mobile apps that use public sector information as a direct input from
data.gov.uk by content category. One might infer that these apps could not have been created in
the absence of these public sector information datasets.
Figure 3.13: data.gov.uk dataset mobile apps, October 2012 106
Source: Deloitte analysis of data.gov.uk data using Google Analytics
What is particularly interesting in Figure 3.14 is that the most viewed datasets on data.gov.uk (as
shown in Figure 3.2) do not entirely correspond with those that are most used to develop apps.
Figure 3.14 compares page views to dataset apps.
105
See http://data.gov.uk/node/add/data-request
This analysis was done in October 2012. The number apps that use data.gov.uk datasets is likely to have risen since
then.
106
71
Market Assessment of Public Sector Information
Figure 3.14: data.gov.uk dataset apps versus page views, October 2012
Source: Deloitte analysis of data.gov.uk data using Google Analytics
Public sector information covering transport is clearly the most popular both in terms of page views
and number of apps developed, with government operations data, personal finance information,
housing, crime and justice, geospatial and education also attracting high levels of attention.
Statistics from data.gov.uk on the page views received by publishers on the site indicate that the
Office for National Statistics has received by far the most attention from users, with over three
times the views of the next most viewed publisher, DCLG. This is only partially a consequence of
the fact that the ONS is the most prolific publisher on data.gov.uk, and suggests that ONS datasets
are particularly interesting to users.
72
Market Assessment of Public Sector Information
Figure 3.15: data.gov.uk most viewed publishers (July 2012 – January 2013) 107
Source: Deloitte analysis of http://data.gov.uk/data/site-usage/publisher
Data.gov.uk statistics on the most downloaded datasets give some indication as to which
information users believe holds the most value and may also indicate where PSIHs should focus
their efforts in improving quality and access. Interestingly, many of the datasets in the list of the top
20 most downloaded fall into one of the areas that this report identifies as areas where public
sector information offers particularly promising potential for generating value (see Chapter 5),
including:

transport, especially for roads – live traffic data, traffic counts, road safety;

health – the GP prescribing data has obviously attracted a high level of interest following the
Mastodon C work;

property – prices and energy efficiency; and

social research – low income students in higher education, social trends, average earnings.
107
“Views” is the number of times a page was loaded in users' browsers. (http://data.gov.uk/data/site-usage/publisher)
73
Market Assessment of Public Sector Information
Figure 3.16: data.gov.uk most downloaded datasets 108
Source: www.data.gov.uk/data/site-usage/dataset
Again, it should be noted that the analysis refers only to open data available on data.gov.uk and
therefore misses public sector information from PSIHs such as the Met Office or Land Registry as
well as non-public sector PSIHs. The review of the Trading Funds in the Public Data Group above
suggests the different ways their data is used and re-used.
Use of public sector information across different sectors of the economy
Another way to consider the public sector information demand side is to explore the use of all types
of open data (public sector information and private sector and individual open data) across different
economic sectors. Building on existing analysis by Deloitte, it is possible to create a data intensity
matrix which compares the use of different types of data by industries in the UK 109 . Figure 3.18
shows this report’s estimates as to which economic sectors most consume different types of open
data. It shows, as might be expected, that businesses in the agricultural, forestry and fishing
sectors are major users and re-users of agricultural open data, but they also use/re-use geospatial, environmental and energy, resources and utilities data.
108
"Downloads" is the number of times a user has clicked to download either an original or cached resource for a
particular dataset. Download information is only available from 2nd December 2012 (http://data.gov.uk/data/siteusage/dataset).
109
This has been developed on the basis of balance sheet analysis of major UK firms.
74
Market Assessment of Public Sector Information
Figure 3.17: Use of data across economic sectors
Source: Deloitte analysis
The x-axis charts the different sectors of the economic and the y-axis different types of public
sector information. The width of each economic sector columns compares the relative
‘consumption’ of open data of each sector. For example, ‘public admin and defence, compulsory
social security’ (i.e. the public sector) consumes over 10 per cent of all data and of this figure, 5 per
cent of its consumption is on transport data, 5 per cent on social conditions data, etc. As Figure
3.17 suggests, the construction, real estate, financial and insurance, public sector and arts,
entertainment and recreation sectors are some of the largest users and re-users of open data.
While this analysis has been carried out on the basis of open data, one might hypothesise that a
similar picture holds for public sector information.
Consumer perceptions of public sector information and unmet demand
In order to gauge consumer perceptions and the unmet consumer demand for public sector
information, Deloitte has undertaken an analysis of publically available requests for information and
data on two data portals: data.gov.uk and the London Datastore. Both these portals allow users to
lodge requests for additional types of public sector information. For a number of reasons, these
results should be treated as indicative only: the sample size is small and inevitably self-selecting. It
is likely that there are many more potential users who have not engaged with the data portals.
Nonetheless, this analysis provides an indicative insight into the types of public sector information
that users are interesting in gaining access to, as well as the uses to which this public sector
information might be put.
75
Market Assessment of Public Sector Information
Data.gov.uk requests
This analysis is based on 100 requests posted publically on data.gov.uk between 26 September
2012 and 18 November 2012. As of the end of November 2012 there were 472 requests posted
publically on data.gov.uk; however, the detail contained in the requests was much reduced after
the first 100. For this reason the 100 most recent requests are used as the sample for this analysis.
Figure 3.18: usage categories cited in data requests (%) 110
Source: Deloitte analysis of data.gov.uk data requests: based on 100 requests
The requests in the sample most commonly cite business use and research as the intended use
for the public sector information, at around 50 per cent for each of these categories.
In terms of the data categories requested, it is location (or geospatial) data that tops the list of
requests. This reinforces the impression that this category of data generates some of the highest
levels of value – or at least that this is where users perceive value to lie. Environment,
transportation, health and society are also data that attract particularly high levels of requests.
Requests for data used to scrutinise and hold government accountable, such as government,
finance, policy, administration and spending data are significantly lower, suggesting that this
attracts less interest. However, even in these categories the numbers of requests are not
insignificant.
110
Note that where the sum of responses shown in the charts is greater than 100, this is because more than one
category is selected in some data requests. For example, a request may cite both business and personal use.
76
Market Assessment of Public Sector Information
Figure 3.19: data categories cited in data requests (%)
Source: Deloitte analysis of data.gov.uk data requests: based on 100 requests
Of the specific barriers cited in accessing public section information, data format and financial
charges are the most commonly mentioned – though these are both behind the ‘other’ category.
Figure 3.20: barriers cited in data requests (%)
Source: Deloitte analysis of data.gov.uk data requests: based on 100 requests
The large majority of requests are generated by private individuals and business, at 41 per cent
and 39 per cent of the sample respectively.
77
Market Assessment of Public Sector Information
Figure 3.21: affiliation of those requesting data (%)
Source: Deloitte analysis of data.gov.uk data requests: based on 100 requests
The most commonly cited data holders are ‘unknown’ and ‘other public body’, suggesting that
users may be unclear where the data they seek is held. Government departments are cited in 19
per cent of requests, although this category covers a large range of PSIHs.
Figure 3.22: data holders cited in requests (%)
Sour
ce: Deloitte analysis of data.gov.uk data requests: based on 100 requests
The Open Data User Group has recently launched a new dashboard on data.gov.uk, providing a
‘roadmap’ of data requests. This promises to make the volume and nature of requests more
transparent and easier to analyse. As the volume of data on requests increases, this should offer
additional insights into the types of data that users and potential users would like to access, and
the barriers they face.
78
Market Assessment of Public Sector Information
Figure 3.23: data.gov.uk request dashboard
Source: www.data.gov.uk/odug-roadmap
London Datastore requests
A similar analysis is present below using data requests from the London Datastore. The sample
size for this analysis is 50 of the data requests with the most ‘wants’ on london.data.gov.uk, out of
a total of at least 870 published requests (as of 14 December 2012). The website allows users to
vote for requests by adding ‘wants’, which function in a similar manner to ‘likes’ on Facebook. The
dates of the requests included in the analysis run from February 2010 to August 2012. The option
to submit a request is currently closed: the website notes: “please note we have temporarily
suspended the dataset suggestion function as a result of recent spamming activity on the site.”
Indeed, the majority of the most recent submissions appear to be spam.
Many of the datasets requested are now available, suggesting that the London Datastore has been
responsive to requests (although it is possible that these requests overlooked already published
data, or that its publication was already planned). Although the sample is small and some of the
requests are over two years old, these requests provide an indicative insight into the sorts of data
to which users and potential users would like to gain access. Note that some of the requests cite
more than one usage category and data category, meaning that in these cases the sum of the
responses is greater than the sample size.
As with data.gov.uk (and reflecting particularities in London), transport is the most commonly
requested type of data, at 40 per cent of requests, with demographic data the second most
common at 22 per cent. This is likely to reflect the high perceived value of these types of data,
particularly on a local level, as witnessed by the rapid increase in the number of mobile apps
available that use Transport for London data.
79
Market Assessment of Public Sector Information
Figure 3.24: most requested data categories
Source: Deloitte analysis of London.data.gov.uk data requests
When the number of ‘wants’ received by request is analysed, transport data is even more popular,
with 50 per cent of ‘wants’ in the sample. This is testament to the perceived value of transport data
to users – although it may also reflect lack of awareness as to the possibilities offered by other,
less obviously valuable types of data.
Figure 3.25: most popular data categories by number of ‘wants’
Source: Deloitte analysis of London.data.gov.uk data requests
Public sector information customer types
While this report has not conducted in-depth primary research to explore the details of how
different customers use and re-use public sector information, it is possible to create seven
archetypes of customers based on the available evidence 111 . These groups are summarised in
Figure 3.26 below.
111
It has not been possible to specify, in most cases, precisely what types of public sector information different types of
customers use and re-use, nor the revenue they generate from this data. Nor, in most cases, is it possible to discern
exactly how businesses use public sector information, i.e. its role as an input into products and services as part of the
value chain. This is largely because users, for reasons of commercial sensitivity, have been unwilling to share such
details.
80
Market Assessment of Public Sector Information
Figure 3.26: public sector information customer archetypes
Source: Deloitte analysis
Market Assessment of Public Sector Information
As noted earlier, almost every aspect of the economy will use public sector information in some
respect (either directly as an input into a production process or indirectly as a final consumer using
products and services built on or around public sector information); it is therefore probably true to
say that these archetypes are not absolutely comprehensive in their coverage of all aspects of
public sector information use and re-use. Nonetheless, based on this report’s survey of information
users, they capture the most important areas where public sector information is used as an input
and generates value.
Large data services company
Companies that offer data services use data from a variety of sources, including public sector
information, to deliver value-added products to clients. They are likely to offer some or all of the
following services:

Credit Services: helping organisations to evaluate the risks and rewards associated with
providing credit to consumers and businesses, enabling clients to make better informed lending
decisions.

Decision Analytics: providing the analytical skills and specialist software products that enable
organisations to increase the speed and quality of their decision-making, helping clients to
optimise their lending strategies and to implement changes quickly.

Marketing Solutions: helping organisations to find new customers and to take advantage of
opportunities for expanding existing relationships, enabling clients to communicate with
prospective customers in the most effective way and with the most appropriate offer.

Interactive Enabling: allowing consumers to view their credit report online and to monitor
changes to their credit records.
Depending on the service offered, a range of types of public sector information is likely to be used
as an input. This includes information on companies (such as supplied by Companies House),
demographic information, and economic forecasts. This information is likely to provide one input
among many, and so only a proportion of revenue could be directly traced back to public sector
information. In many cases, the products being sold by the large data services company could still
be produced, albeit not as effective. In isolated cases, substitutes may also be found to the public
sector information, albeit potentially at greater cost or reduced quality.
Independent or SME app developer
App developers are predominately individuals or small companies 112 with the technical skills to
create smartphone applications presenting publically available data, including public sector
information, to a wider base of non-technical users. They may release these apps free of charge,
or for payment; or they may release multiple versions, i.e. a free version to download but
containing advertising, and the other requiring payment but no advertisements. An example is the
developer Routemaster, responsible for releasing the London Transport Live and London
Transport Pro apps using Transport for London data. 113
The apps market as a whole is fragmented among many individuals and small enterprises and
individual profitability is not easily known. Nonetheless they are an important conduit to a wider
range of users and re-users for public sector information.
The type of public sector information used varies depending on the nature of the app. For example,
transport apps will use road and rail data; weather forecasting apps will use data from the Met
Office; property price apps will use data from the Land Registry; etc. In many cases there will be no
substitute for public sector information as an input, or any substitute may be prohibitively priced or
of lower quality. High quality, reliable and competitively priced or free public sector information
112
113
Though of course, larger companies will also produce apps, but not as their sole activity.
See https://play.google.com/store/apps/developer?id=RouteMaster&hl=en_GB
Market Assessment of Public Sector Information
therefore underpins many app products, and is likely to play an increasing role as more public
sector information become available and the app market matures.
SME focussed on policy efficiency solutions
SMEs using public sector information and other information sources to develop products that aid
policy efficiency can usefully be grouped into a third archetype.
An example is ELGIN, which via its service roadworks.org publishes streetworks and other
highway information on the web. It is designed to facilitate coordination of activities between
neighbouring authorities and statutory service providers, such as utility companies, enabling them
to reduce road space occupation and to meet their statutory obligations under the Traffic
Management Act. 114
The ELGIN business model is a good example of a shared geospatial ‘cloud-based’ service,
where, for a relatively small upfront subscription, joining authorities get access to shared services
and obviate developing and supporting their own computer system. The approach also supports
interoperability and data sharing, with all information being presented in a common format on a
single website.
In this instance public sector information (transport information, in this case) is critical to the
business model of roadworks.org, without which it could not exist. This is therefore an example of a
business that has been enabled through the release of public sector information, and which acts as
an ‘enabler’ to facilitate improved use of this information (see below).
In many cases, business models may rely on combining public sector information with other
sources of information. For example, Honest Buildings intends to combine public sector information
such as the Energy Performance Certificates database (environmental/energy information) with
other sources of information on properties and real estate services, helping property owners to
realise energy savings through improvements to their properties. 115 In these instances the
business could often exist without the public sector information, but the value of its offering is
greatly enhanced by access to this information. This is therefore an example of an area where
government could add value and aid the development of new markets through targeted release of
key public sector information datasets (the Energy Performance Certificates database being a key
current example).
Individual users
Individuals with the technical skills and interest to work with data may decide to use their skills to
derive insights from data (research or other purposes) or present it in a form that is more
accessible to non-technical users. For example, upon the release of road construction project data
by the Department of Transport in Edmonton, Canada, a local application developer decided to
create a mobile app for smart phones and similar devices to access the map interface. He saw the
data as useful for the population and decided to make a contribution to his new home city, not as a
commercial venture. He reported that creating the app was a relatively straightforward task for him,
due in part to his extensive development experience and the high quality of the data set and
metadata provided by the city, indicating that these individuals often lack the resources to exploit
data if it is not presented in a suitable format. 116
In these cases, the monetary value generated by the use of public sector information may be small
or non-existent, but the product may produce wider social and economic benefits. In general,
availability of public sector information appears to be a precondition for the emergence of many of
these-user generated products, as alternative sources of information may not exist or may be
prohibitively priced. However, this varies on a case by case basis.
114
See http://roadworks.org/
See http://www.honestbuildings.com/
116
Centre for Technology in Government, 2012. The Dynamics of Opening Government Data: A White Paper.
http://www.ctg.albany.edu/news/press_ogsap2_20121204
115
83
Market Assessment of Public Sector Information
Not for profit organisations
Not for profit organisations may use public sector information for a wide range of purposes,
including holding elected officials to account, strengthening civil society, and providing means for
citizens to monitor and influence areas of interest to them such as the environment and local
services. As such public sector information use or re-use by these organisations does not
necessarily have a revenue impact, but is likely to have wider social, economic or public interest
benefits. Their task would in many cases be impossible to perform without access to public sector
information. The benefits arising from their actions can therefore in many cases be largely
attributed to the availability of public sector information.
General data user (non-specialist company)
Many companies, notably professional services firms, use public sector information as a matter of
course as a key source to inform their work. An example of public sector information commonly
used by such companies is ONS data on the economy. The conduct of their operations would
become more difficult and more expensive were public sector information unavailable. However,
while these firms work with public sector information and other data on a day to day business, they
differ from data services companies in that they do not explicitly sell data services. The proportion
of their revenue attributable to public sector information is therefore likely to be lower than for the
data service companies, and it is likely that they could turn to alternative sources of information
were the public sector information unavailable. In these cases public sector information generally
improves the product or service offered, rather than being integral to its existence.
Public sector bodies
Public sector bodies not only produce but also use public sector information in their day to day
operations. Uses vary widely, ranging from use of address data by local authorities to use of
economic data by central government departments. Public sector information use by public sector
bodies does not necessarily generate revenue but may assist with achieving more efficient and
effective operations, leading to cost savings and improved outcomes.
The ways in which these different archetypes generate value from public sector information is
shown below. Value can be generated through its independent use or re-use or by combining
public sector information with other data sources to become important inputs into products,
services, decision-making and other outcomes. As noted above, public sector information acts as
an input as a data point for complex algorithms and analyses, as a source of insights (perhaps
summarised as data dashboards or visualisations), as API feeds for apps and so forth.
84
Market Assessment of Public Sector Information
Figure 3.27: public sector information value chain
Source: Deloitte analysis
The following section focused on a particular group of business models used by public sector
information intermediaries, who can be found in the SME and large data service companies
archetypes.
85
Market Assessment of Public Sector Information
Different business models of public sector information intermediaries
The ways in which different intermediaries use and re-use public sector information varies by their
business models. To recap, the public sector information value chain includes suppliers (PSIHs),
intermediaries (enablers and infomediaries) and final consumers. Figure 3.28 summarises the
different business models currently in operation across the public sector information market.
Figure 3.28: public sector information and open data business models
Source: Deloitte analysis
As Figure 3.28 shows, there is clearly overlap in business models that rely upon open data more
generally with those that rely specifically on public sector information (either as open data or
otherwise). The business models here are a way of deriving economic benefit out of public sector
information and open data. Many of these might be interrelated, such as, in some cases, Open API
is the data feed for Apps based reuse, and hence an Enabler. The following sections briefly outline
each business model, as well as the economic impact arising from each (the relationship between
economic growth and public sector information and open data is explored in much more detail in
the next chapter).
PSIHs as public sector information suppliers – business models
The collection, retention and distribution of public sector information is not without cost – though
the level of cost will vary by dataset and PSIH. In some cases, PSIHs will seek to recover these
costs from intermediates and final consumers – indeed, they are often incentivised to do so in
terms of legislative requirements or other business objectives. The different ways in which public
sector information can be monetised are shown below.
Figure 3.29: public sector information monetisation models
Source: Adapted from Programmerweb, http://programmersweb.blogspot.co.uk/
86
Market Assessment of Public Sector Information
The choice of which monetisation model to use will depend on each PSIH’s strategic objectives
and the type of public sector information in question. For example, Trading Funds are explicitly
required to make a return on data collected, retained and disseminated.
There are a number of benefits to releasing public sector information for PSIHs regardless of the
monetisation model. Through open application programming interfaces (APIs) PSIHs can improve
their brand visibility, explore new distribution channels, extend innovative external product
development, and tap into new communities.
Intermediaries as data marketplaces – business models
Intermediaries acting as data marketplaces or infomediaries have very successfully monetised
public sector information and other open data as a third-party reuse, mostly through freemium
model or advertisement revenues. Figure 3.30 provides some examples of different business
models.
Figure 3.30: data marketplaces business models
Website
Data Domain
Monetisation
Factual
High quality local data
Travel, finance, sports, autos,
movies, music, TV, books,
health, food, politics,
education, science, arts
Currently free
Pay-per-use pricing for every
API call with subscription
Integrate third party deals in
apps
Open data sources,
community
contribution, web
crawling
Infochimps
General public sector
information
Charge data sellers
Charge data buyers
User submitted data
Data from public
sources
Datamarket
Public sector information
Statistical data
Charge data sellers
Charge data buyers
Datasets from public
sources
User submitted
datasets
Freebase
General public sector
information
Paid for higher volumes of
data calls
Community curated
data
Public datasets
imports
Kasabi
All-purpose public sector
information, including BBC
linked data, Geonames
Charge data consumers
Public datasets
User submitted
datasets
Timetric
Public sector information
economic data
Free public datasets
Paid exclusive datasets
Public sources of
economic data
User uploaded
datasets
Paid subscription
Aggregate data from
leading financial data
sources, public
datasets, and user
uploaded data
Xlgnite
Financial data
Sources
Source: Deloitte analysis
Chapter Five discusses the ways in which the use and re-use of public sector information can
generate impact, but focusing just on data marketplaces and infomediaries, one can see how other
businesses, individuals and the public sector benefit from their products and services - not just the
data marketplaces and infomediaries themselves, through increased profits.
87
Market Assessment of Public Sector Information
Figure 3.31: impact of data marketplaces
Source: Deloitte analysis
Apps based re-use – business models
App developers take advantage of refined public sector information and other open data processed
by infomediaries and enablers to produce apps for end-users – either for their smart phones,
tablets, laptops or other devices. The barriers to entry into the app development market are very
low and there are a range of established distribution channels available.
Some apps are made available for free, while others are available upon a fee or on a freemium
basis. Consumers’ willingness to pay for apps will ultimately depend on their expected utility from
using these apps. For example, though people can access The Guardian as a mobile site on their
iPhone, The Guardian iPhone app is among the top paid application. Similarly, citizens in San
Francisco pay for more reliable and more interactive iBARTLive (Bay Area Rapid Transit System),
although it is available as a mobile site, and is integrated with Google Maps. It may also be that
willingness to pay for apps exceeds willingness to pay for websites, owing to the greater
convenience.
Figure 3.32: app value-chain
Source: Deloitte analysis
Figure 3.33 above shows the value-chain for app development from public sector information and
other open data. Clearly, there are significant benefits to be had from app-based public sector
information and open data re-use. Beginning with narrow effects, businesses creating apps can
create new jobs and generate wealth. This in turn can generate supply-chain or indirect impacts:
changes in the number of jobs and income in associated industries that supply inputs to
88
Market Assessment of Public Sector Information
organisations developing apps. Finally, there will be induced or consumer impacts through
spending by households associated with businesses creating apps and the supply chain that
results in changes to jobs and income. Additionally, there may be broad effects such as apps
facilitating greater participation in policymaking or more informed choices.
Figure 3.33: impact of app re-use
Source: Deloitte analysis
Public sector information and open data backed 'Apps' have already reduced fact-checking
costs/time considerably, and led to increase in citizen awareness and partnership, with the case
study below providing an example of this.
Figure 3.34: London Olympic Games – public sector information app use
The proportion of the UK population using smart phones has risen from 30 per cent in 2010 to 44 per
cent in 2012 – it is estimated 4 in 10 adults use their phone to go online.
The presentation of London Olympic Games-related public information in app format led to more
usage than the mobile web – indicating that citizens found this more useful accessing information via a
webpage.
Source: Adult Media Use and Attitude Report, 2012, Ofcom; Neilson
Data enrichment – business models
Data enrichment refers to those consumers who take advantage of public sector information to
improve and expand their existing products and services. In its broadest sense this could be
improved decision-making with reference to the latest economic and demographic statistics
published by the ONS, or more specifically, it could be the embedding of newly available
89
Market Assessment of Public Sector Information
educational datasets to refine existing educational products focusing on those features that truly
improve outcomes.
The types of business which ‘enrich’ data and their end consumers are summarised below in
Figure 3.35.
Figure 3.35: data enrichers
Source: Deloitte analysis
As can be seen from Figure 3.35, data enrichers can be found across all aspects of the economy.
Data enablers – business models
The final category of business model this report has identified is data enablers, who are another
group of intermediaries. Data enablers are businesses that facilitate the public sector information
and open data environment through:

data storage infrastructure;

platform, hosts, network provider;

consulting /advisory services;

software provision;

providing devices where apps can be accessed, - mobile phone companies, operating
platforms provider; and

the developer community.
Through their work, enablers allow other businesses, government and individuals to more readily
exploit public sector information.
90
Market Assessment of Public Sector Information
Figure 3.36: data enablers
Source: Deloitte analysis
Data scientists
Individuals using and re-using public sector information on behalf of their organisation or for
personal use are likely to have a range of analytical skills and abilities. The types of roles that are
developing around information and data exploitation are well demonstrated by the example of the
emerging role of the ‘data scientist’.
Figure 3.37: profession profile: data scientists
A key profession in the world of open data, public sector information and data analytics is the ‘data
scientist’, which was recently described by The Harvard Business Review as “the sexiest job of the
21st century”. The data scientist is described as “a high-ranking professional with the training and
curiosity to make discoveries in the world of big data.”
Although statisticians and experts on quantitative analysis have long existed, data scientists differ from
these existing professions in a number of important ways. As well as being able to work with large
volumes of structured and unstructured data, they are able to translate these analyses into policy and
commercial-ready insights and effectively communicate them to a range of stakeholders, often using
innovative tools and visualisations. Key to this is the ability to “identify rich data sources, join them with
other, potentially incomplete data sources, and clean the resulting set.” In many ways they therefore
resemble scientists more closely than traditional data analysts.
Data scientists are generally drawn from the ranks of graduates in numerate and scientific fields,
including mathematics, economics, and computer science. Scientific qualifications with a less
immediately obvious relevance to data science, such as ecology and astrophysics, have also
generated successful data scientists, as many of the core traits of practitioners in these fields are
similar: an ability to generate and test hypotheses and the curiosity to penetrate to the heart of a
problem.
The data scientist is a best thought of as a “hybrid of data hacker, analyst, communicator, and trusted
adviser.” Keys skills required for the role of data scientist include:

The ability to write computer code

A foundation in maths, statistics and probability

Skill in communicating the ‘story’ the data tells to a non-technical audience

Ability to display complex information visually
91
Market Assessment of Public Sector Information
The key traits of a typical data scientist may include:

Curiosity

Empathy and a feel for the key issues

Creativity and imagination
Source: Harvard Business Review, ‘Data Scientist: The Sexiest Job of the 21st Century’ (October 2012);
Examples of businesses directly employing data scientists include:

GE uses data science to optimize the service contracts and maintenance intervals for industrial
products;

Google uses data scientists to refine its core search and ad-serving algorithms;

Zynga, a games company, uses data scientists to optimize the game experience for both longterm engagement and revenue; and

Kaplan, a test preparation company, uses its data scientists to uncover effective learning
strategies.
Based on ONS figures for 2011, around 1.5 million workers in the UK, representing around five per
cent of the active workforce, are employed in job categories that are likely to involve elements of
the role of the data scientist 117 . These figures are indicative only, owing to limitations with the ONS
data and the use of SOC occupational classifications, but indicate how widespread the requirement
for data analysis skills has become. The importance of these workers to the UK economy is likely
to be greater than the numbers indicate as they are usually employed in high-skilled, highproductivity, and high-paid job categories.
A recent report 118 has estimated current and future demand for staff with data skills. It found that
data scientists, while important, accounted for less than 1 per cent of all advertised positions for big
data staff in the third quarter of 2012. However, when the analysis expanded to consider other
roles not labelled data scientists, but likely to perform related roles (i.e. developers, IT architects
and analysts), these were among some of the most commonly advertised roles.
Figure 3.38: most commonly advertised big data roles, Q3 2012 (% of total)
Source: ‘Big Data Analytics’ (January 2013)
The report found an observable pay premium for big data staff, with salaries around 20 per cent
higher than those for IT staff as a whole. This reflects the rapid increase in demand for big data
117
These categories include: science, engineering and technology professionals, IT professionals, statisticians,
economists, management consultants, actuaries, research managers and architects and systems designers.
118
e-skills UK, ‘Big Data Analytics: an assessment of demand for labour and skills, 2012-2017’ (January 2013)
92
Market Assessment of Public Sector Information
staff and a potential shortage of skills: the report found that demand has risen by 912 per cent over
the last five years. This includes a 350 per cent rise in demand for data scientists. This growth is
projected to continue: depending on the growth scenario adopted, the report anticipates growth in
demand for big data staff of between 13 per and 23 per cent per annum. Taking a midpoint of 18
per cent would mean the generation of 28,000 gross job opportunities per annum by 2017, with
132,000 gross job opportunities created between 2012 and 2017.
Analysis of ONS SOC occupations that involve an element of data science for this report found that
the average annual median wage of these ‘data science’ workers was over £36,000 in 2011 which
compared favourably to the national annual median wage of £26,000 in 2011.
Chapter summary

While data on the total number of public sector information datasets is not recorded, a review of selected
data portals suggests the number could exceed 37,000.

Some of the key suppliers of public sector information, in terms of number of datasets, include the Office for
National Statistics, the Department for Communities and Local Government and the Health and Social Care
Information Centre. Collectively, the top ten publishers of public sector information make up 60 per cent of
all datasets on data.gov.uk.

Datasets on government spending are among the most available datasets.

Levels of awareness of public sector information vary across staff in PSIHs. Over 50 per cent of civil and
public servants, according to one survey, did not know how to access datasets, how to interpret them or
best apply data standards.

The demand side of the market includes infomediaries and final consumers. Business models include
developing apps that use/re-use public sector information, data enrichers, data enablers and data
marketplaces.

Building on existing analyses, it is possible to create a data intensity matrix to compare how different types
of datasets are used by different economic sectors in the UK currently. The analysis also shows that the
construction, real estate, financial and insurance, public sector and arts, entertainment and recreation
sectors are some of the largest users and re-users of open data.

Dataset types most commonly being used include demographics, economic, environmental and geo-spatial
and housing data.

From an analysis of data.gov.uk dataset requests, the most popular dataset category requested is location
(or geospatial) data. This reinforces the impression that this category of data generates some of the highest
levels of value. Environment, transportation, health and society are also data that attract particularly high
levels of requests.
93
Market Assessment of Public Sector Information

As well as being used for commercial and policy purposes, public sector information can be used for
scientific research, data journalism and holding public service providers to account.

Some of the most viewed and requested datasets relate to geospatial and transport information. Of the
named barriers identified as preventing access to public sector information, data format and financial
charges were among the most commonly mentioned.
94
4. Public sector information
regulatory landscape
This chapter provides an overview of the key stakeholders involved in
the governance and policy direction of public sector information, as
well as the main regulations and legislation influencing its delivery. It is
not intended to be an exhaustive survey of the regulatory landscape but
provides a high level overview of the salient points.
Policy and regulatory stakeholders
The public sector information market is complex and evolving. The market also continues to be the
subject of a number of policy and other regulatory and legislative initiatives. Many of the policy
questions around public sector information link to wider agendas on transparency, accountability
and growth - within which there are a range of initiatives. Some of these currently include:

a number of consultations covering a range of issues from Freedom of Information, data
release guidelines, the right to data and intellectual property;

a number of forums considering public sector information and transparency issues including
the Data Strategy Board, Transparency Board, the Public Data Group, the Open Data
User Group, the Administrative Data Taskforce, the Advisory Panel on Public Sector
Information and others; and

updating existing policies on data and transparency for bodies such as the Research Councils.
A full list of contemporary initiatives in this area can be found on the Cabinet Office website 119 .
These initiatives notwithstanding, it is possible to sketch out a provisional schematic for the market
that captures the different players and their roles and responsibilities. As well as PSIHs, there are a
range of actors involved in the governance of public sector information and open data policy and
regulations – all of which have a bearing on how the market develops.
The highly simplified schematic in Figure 4.1 illustrates the number of different policymakers and
others involved in shaping the public sector information landscape. Within each PSIH there will be
staff directly involved in public sector information policymaking and others more focused on
delivery. In a number of cases, policy shapers will also be part of the public sector information
supporting infrastructure and may also be public sector information suppliers and consumers.
119
See www.gov.uk/government/policies/improving-the-transparency-and-accountability-of-government-and-its-services
Market Assessment of Public Sector Information
Figure 4.1: simplified public sector information market schematic (not definitive)
(March 2013)
Source: Deloitte analysis. * Some policy shapers will also have operational duties, e.g. Cabinet Office is responsible for data.gov.uk
As the market evolves, one might expect increasing numbers of supporting public sector
information agencies to emerge, as well as organisations in the private and voluntary sectors, as in
the case of the Open Data Institute.
The legal framework for publishing and sharing public sector
information120
The ability or inability of government departments and other PSIHs to publish and share public
sector information they hold is dependent on their possessing one or more of the following:

express statutory powers: these are powers specifically given through legislation;

implied statutory powers: these are powers not specifically given in legislation but which can
be implied from it because without these powers the public body would be unable to carry out
its functions as specified in legislation; or

common law powers: these are powers that exist outside of legislation. The Ram Doctrine
(after Sir Greville Ram, 1945) states that a Crown government department has all the powers
of a natural person and does not need to demonstrate any statutory power or authority to carry
out an action, unless that action is expressly or implicitly precluded by statute 121 .
120
Note, this section relies heavily on the report by the Administrative Data Taskforce in December 2012: Improving
Access for Research and Policy. This is available at: www.esrc.ac.uk/_images/ADT-Improving-Access-for-Research-andPolicy_tcm8-24462.pdf.
121
Administrative Data Taskforce, ‘The UK Administrative Data Network: Improving Access for Research and Policy’
(December 2012) (ADT)
96
Market Assessment of Public Sector Information
There are notable cases where statutory powers restrict or supersede common law powers. For
example, the Department of Education shares individual pupil data under the Education Act 1996,
which sets the limits for sharing this data. 122 However, it continues to share some of its aggregated
data under common law powers. Consequently there are statutory restrictions on the sharing of
individual pupil data which do not necessarily apply to the sharing of the same data in aggregate
form. This level of complexity means both that it is not always clear whether existing legal powers
(statutory and common law) are sufficient to allow a particular instance of data sharing; and also
that even where there is a legal basis to permit sharing or release of data, there can be uncertainty
over the legal position leading to public sector bodies ‘playing it safe’ and restricting data sharing.
Given that unlawfully sharing information relating to an identifiable person or legal entity is an
offence that can carry a prison sentence of up to two years, 123 this caution is understandable.
Nonetheless, it can create additional barriers to sharing of information even where it is permitted
under statute or common law.
Where two or more public sector bodies attempt to share information, the level of complexity is
even greater. It may be that one body is legally permitted to share its information but the other is
not. In this case the second body may be able to carry out the linking work, but this still raises the
question of whether the resulting analysis can legally be shared with the first body. The level of
complexity only increases as the number of public bodies involved rises – and many of the most
interesting opportunities for linked data may involve information held across the full spectrum of the
public sector.
In specific cases a ‘legal gateway’ may be created through legislation in order to explicitly allow the
sharing of information. An example is data collected under the Statistic of Trade Act 1947, which
may be provided to local authorities owing to subsequently created legal gateways, the
Employment and Training Act 1973, as amended by the Employment Act 1988. These allow the
sharing of relevant information with the persons listed under section 4(3) of the 1973 Act.
However, as is indicated by the above example, these legal gateways are complex and timeconsuming to set up – it has been reported it can take as long as two years, by which time the
window of opportunity to enhance a policy through sharing information may well have passed.
Having been created, the legal gateway only applies to a very specific instance of information
sharing, while the proliferation of such gateways continually increases the legal and regulatory
complexity of the public sector information landscape. In order to be clear as to whether a
particular instance of information sharing is permitted, an employee of a public sector body
requires a substantial knowledge and understanding of legally complex gateway arrangements.
In its recent report the Administrative Data Taskforce recommended the creation of “a generic legal
gateway for research access to and linkage between administrative data.” 124 The rationale is to
create a more consistent, efficient and timely legal framework for sharing information, so as to
enhance policymaking with evidence-based insights and to ensure that information holders have
clear guidance as to when it is appropriate to share information.
Public sector information licensing and the regulatory framework
The UK Government Licensing Framework
Public sector information provision is governed by the UK Government Licensing Framework
(UKGLF), overseen by The National Archives. It provides “a policy and legal overview of the
arrangements for licensing the use and re-use of public sector information both in central
122
At the time of writing, the DfE was consulting on new access arrangements to the National Pupil Database.
123
ADT (2012), p. 16
124
ADT (2012)
97
Market Assessment of Public Sector Information
government and the wider public sector.” 125 The key licences comprising the UKGLF are
summarised below.
Open Government Licence
The Open Government Licence (OGL) was introduced in 2010 as a licensing mechanism to facilitate the use
and re-use of public sector information. It is designed to impose as few conditions as possible on users,
permitting them to:

Copy, publish, distribute and transmit the information;

Adapt the information; and

Exploit the information commercially, including by combining it with other information and including it in a
product or service.
Notably, information published under the OGL does not require users and re-users to register or pay any
charge.
The main condition of use for information published under the OGL is that users must acknowledge the source
of the information by providing any attribution statement specified by the information provider.
Other conditions and limitations are that:

Users may not use the information in a way that suggests official status, or that they have the
endorsement of the information provider;

Users may not mislead others, nor misrepresent the information or the information provider; and

Users must ensure that their use of the information does not breach the Data Protection Act 1998 or the
Privacy and Electronic Communications Regulations 2003 (EC Directive).
The OGL also limits the liability of the information provider by specifying that the provider is not liable for any
errors or omissions in the information, and does not guarantee the continued supply of the information.
The OGL replaces the previous Click-Use Licence as the default licence for a wide range of information owned
by the Crown, including information previously made available under the Click-Use Licence and source code
and software originated by the Crown under Framework 1 of the NESTA agreements. It is designed to be
interoperable with other widely used models including the Creative Commons Attribution Licence and the Open
Data Commons Attribution Licence. It is increasingly widely used across the entire public sector, and its use is
not confined to Crown bodies. For example, over 270 local authorities currently use the OGL.
Source: The National Archives
Non-Commercial Government Licence
Although the default position is that information should be published under the OGL, there are circumstances
under which it is not appropriate to allow commercial use of information. In these circumstances information
providers can publish under the Non-Commercial Government Licence.
This licence allows the use and re-use of information free of charge, permitting users to:

Copy, publish, distribute and transmit the information;

Adapt the information; and
 Combine the information with other information.
The main restriction on the use of information published under this licence is that it may not be used for
purposes intended to confer commercial advantage or private monetary compensation. These restrictions also
apply to any onward licensing of the information, for example when combined with other information. As with
the OGL, users are required to acknowledge the source of the information.
Source: The National Archives
125
See www.nationalarchives.gov.uk/information-management/government-licensing/the-framework.htm
98
Market Assessment of Public Sector Information
In certain circumstances public sector bodies are permitted to charge for use and re-use of
information on the basis of cost recovery and a reasonable rate of return. The most notable
examples of this are the Trading Funds, although they are not the only examples.
In such cases, the specific terms of the licence vary depending on the requirements for the product
or service in question. However, the National Archives is in the process of developing a standard
charged licence under the UKGLF, currently in beta version and undergoing development at the
time of writing. The aim is to “provide a straightforward set of terms which deliver an effective
standard approach.” 126 This is intended to provide greater consistency and transparency around
licensing conditions where information is charged for, both for the purchaser and the provider of
the information, and thereby reduce costs for the provider and increase the accessibility of the
information for users.
An additional licence is the Open Parliament Licence, which is designed to enable parliamentary
material to be shared and used in a manner consistent with the principles of the OGL. 127
Crown copyright and the Information Fair Trader Scheme
The National Archives manages Crown copyright by licensing material, and by granting
delegations of authority to certain public sector bodies for certain information to allow them to
license the material they hold. Delegations are only offered to certain organisations, notably trading
funds; other organisations must write a business case setting out why this is appropriate.
This is regulated through the Information Fair Trader Scheme (IFTS), which is designed to ensure
that re-users of information are treated fairly by public sector information providers. IFTS
accreditation is based on a full audit of information trading activities and is intended to show that
“their processes and policies are compliant and consistent with government policy on information
trading and that they meet the needs of existing or potential customers.” All public sector bodies
that license information can apply to join the IFTS. Organisations currently accredited include Land
Registry, Ordnance Survey, the Met Office and Companies House, among other trading funds. 128
The IFTS is based on the following principles: 129

Maximisation: the default position should be that information can be re-used, in the absence
of strong reasons to the contrary;

Simplicity: regarding processes, policies and licences;

Innovation: actively facilitate the development of new and innovative forms of re-use;

Transparency: regarding terms of re-using, charging details, and what information is available;

Fairness: re-users should be treated in a non-discriminatory way, and information holders
should not use their position to compete unfairly; and

Challenge: a robust complaints process.
Exceptions to marginal cost pricing
As discussed above, most public sector information is made available at marginal cost, which in
practice usually means free of charge, especially where the information is provided online.
However, there are exceptions to this. Where a PSIH wishes to charge above marginal cost for
information, it must make a business case to the Office of Public Sector Information 130 under the
126
See www.nationalarchives.gov.uk/information-management/government-licensing/charged-licence.htm
See www.parliament.uk/site-information/copyright/
128
See www.nationalarchives.gov.uk/information-management/ifts/full-accreditation.htm
129
See www.nationalarchives.gov.uk/information-management/ifts/principles.htm
130
The Office of Public Sector Information operates from within The National Archives and has a statutory role under the
PSI Regulations for the investigation of complaints.
127
99
Market Assessment of Public Sector Information
exceptions process. An application will generally be made if an information holder has added value
to the information and therefore wishes to treat it as a commercial product.
The case for an exception is judged on the following criteria: 131

Is it essential to produce the information as part of government’s core duties and therefore vital
to the workings of government?

Is the information directly funded by the taxpayer, either through it being collected for the
purposes of government or produced with the purpose of informing the public?

Is the Department or Agency the sole producer of the information or can the information be
obtained from other sources?

What effect would charging for the information have on the level of re-use?

Is the Department or Agency able to provide a statement of commitment to Information Fair
Trader principles signed by its Permanent Secretary or Chief Executive?
These criteria are designed to ensure that charging above marginal cost does not compromise the
core principles of public sector information policy, especially:

maximisation of access, re-use and innovation;

openness and accountability; and

competitive neutrality.
The below schematic illustrates the exceptions application process.
131
See www.nationalarchives.gov.uk/documents/information-management/criteria-exceptions-marginal-cost-pricing.pdf
100
Market Assessment of Public Sector Information
Figure 4.2: the exceptions application process
Source: adapted from www.nationalarchives.gov.uk/documents/information-management/how-we-deal-with-exception-application.pdf
The Data Protection Act
The Data Protection Act 1998 (DPA) was enacted to bring UK law into line with the EU Data
Protection Directive 1995, which required Member States to protect citizens’ rights and freedoms,
including their right to privacy, with regard to the processing of personal data. As such it plays a
significant role in the public sector information landscape, particularly with regard to sharing and
publishing data which may have personal privacy implications.
The DPA covers any data which concerns a living and identifiable individual. Anonymised or
pseudonimised data is not considered personal data and is therefore not covered by the DPA – in
this respect the UK differs from several other EU Member States. Where anonymisation is
reversible, however, the data does fall within the scope of the DPA.
The individual who is the subject of the data that is held or processed is entitled to:

view the data for a ‘subject access fee’;

request that incorrect information be corrected;

require that data is not used in any way that could cause damage or distress; and

require that their data is not used for direct marketing purposes.
The subject of the data must consent to the collection and storage of their information. The data
holder has an obligation to inform the data subject the purposes for which it is being used and to
whom it has been disclosed, and the data subject may be able to withdraw consent. In any event,
101
Market Assessment of Public Sector Information
consent is not assumed to last forever – in general it is assumed to last for as long as the data
needs to be processed.
In some cases the DPA may be seen as a barrier to the release or sharing of information, or it may
be that information holders are unsure of their obligations and therefore err on the side of caution
and resist releasing information even where this is permitted. However, from conversations with
stakeholders in the information and regulatory communities, notably The National Archives and the
Information Commissioner’s Office, it seems clear that while the DPA requires information holders
to handle information with care: it should not be a barrier to its use and re-use. For example, if the
information undergoes appropriate anonymisation its use and re-use is no longer constrained by
the DPA, because anonymised information is not considered personal information.
It should be noted that the reform of the EU legal framework on the protection of personal data,
proposed by the European Commission in 2012 and currently undergoing negotiation, may
eventually replace the DPA with a new EU regulation. The impact on the UK regulatory landscape
is currently unclear since the details of any new regulation are as yet unknown.
The case study below is an example of the anonymisation process to which data can be subjected,
in order to ensure that it does not contravene the principles of personal privacy or fall within the
remit of the DPA.
Case study: avoiding the ‘jigsaw effect’
Anonymisation of datasets is a key plank of the transparency agenda. It is particularly relevant where datasets
are of high value for researchers, policymakers and the wider public, but also contain personal and potentially
confidential details of individual citizens. Publication of such personal details would be a breach of privacy. In
order to make these datasets available without compromising privacy, key individual identifiers (names, dates
of birth, addresses and so forth) are removed prior to release. High profile examples of these types of datasets
include the National Pupil Database, court sentencing data, and data on offenders and reoffenders.
However, it is sometimes possible to ‘de-anonymise’ data through a process called ‘jigsaw identification’. This
involves assembling information from other sources which allows re-identification of an individual through a
process of triangulation. This process becomes easier the more information is readily available, which is a
matter of concern given the ever-expanding volumes of information available through the Internet, especially
with the rise of social media and indeed the growth of government transparency programmes. In addition, the
growing level of computational power at the disposal of individuals brings the resources required within the
reach of an ever greater number of people.
The authors of a recent paper 132 therefore describe ‘best practice’ for the anonymisation of sensitive datasets,
based on the experience of the Ministry of Justice in anonymising its data on reoffending. As a first step, the
data was anonymised by Ministry of Justice statisticians, who aggregated the data into ranges for each
characteristic as follows: gender, age, offence, establishment/trust, previous offences, whether reoffended and
number of re-offences.
This data was then passed to academics and postgraduate students with relevant experience at the LSE,
Royal Holloway and Southampton University. They were asked to identify the anonymised subjects.
One disclosure resulted from this process, out of over 200,000 cases. This was possible thanks to the profile of
an offender named on a local news site. In response to this, the MoJ statisticians further aggregated the data
by removing information about the offence committed. The data was then passed to data security specialists
Detica, where a team confirmed that the data was secure against the jigsaw effect.
By involving three diverse groups of testers, the effectiveness of the anonymisation process was tested more
robustly and the MoJ was able to release the anonymised data with high level of confidence that individual
privacy had been respected. The involvement of the students was seen as a particularly important element of
the test, as their approach resembled that of hackers and they subjected the data to unorthodox approaches,
highlighting any unforeseen risks or weaknesses in the anonymisation process.
As a consequence, the MoJ statisticians acquired more detailed knowledge of the properties of the data a
greater understanding of the level of aggregation required to preserve the anonymity of the subjects.
132
Kieron O’Hara, Edgar Whitley and Philip Whittal, ‘Avoiding the jigsaw effect: experience with Ministry of Justice
reoffending data’ (December 2011)
102
Market Assessment of Public Sector Information
Case study: avoiding the ‘jigsaw effect’
The authors suggest that this method could usefully be applied across government, as a way of ensuring that
transparency is accompanied by robust methods and a respect for privacy in which the public can have
confidence.
Source: Kieron O’Hara, Edgar Whitley and Philip Whittal, ‘Avoiding the jigsaw effect: experience with Ministry
of Justice reoffending data’ (December 2011)
The European dimension
In addition to UK statute and common law, the European Union (EU) also plays a significant role in
the legal and regulatory framework for public sector information. The main relevant regulations are
the European Union Directive on the Re-Use of Public Sector Information (the ‘PSI Directive’), and
the European Union Data Protection Directive (see box-outs below). It should be noted that both of
these are undergoing revision, with the revised versions likely to be adopted during the course of
2013. These debates are ongoing, and whichever of these dynamics becomes the dominant trend
at the European level may have a significant effect on the development of the public sector
information landscape in the UK.
The European Directive on the Re-Use of Public Sector Information
The breakout box below summarises the current PSI Directive and the proposals for its revision.
The European Directive on the Re-use of Public Sector Information and the proposed
changes to it
In 2003 the European Union introduced the Directive on Re-use of Public Sector Information (2003/98/EC)
which introduced a common legislative framework regulating how public bodies should make their information
available for re-use, including removing barriers and improving transparency. This required public bodies to:

be transparent on conditions of re-use;

avoid any form of discrimination between users, including a re-use by the public sector body itself;

deal with applications for re-use within a set maximum time; and
 not enter into exclusive arrangements, except under exceptional circumstances.
In December 2011 the European Commission, as part of an Open Data package, presented proposals to
revise the 2003 Directive. 133 These revisions are likely to create a presumption of openness on public sector
information, with key changes, according to the EC, including:

bringing new bodies under the scope of the Directive, including cultural bodies such as museums,
libraries and archives;

limiting the fees that can be charged by public authorities at marginal cost as a rule;

introducing independent oversight over re-use rules in member states; and
 making machine-readable formats the norm for information held by public authorities
Negotiations are continuing at the time of writing.
The revised Directive is expected to be adopted in Europe in the course of 2013 and come into force through
regulations in UK law in 2014-15.
The proposed European Data Protection Regulation
The breakout box below summarises the proposed regulation on data protection.
133
See http://ec.europa.eu/information_society/policy/psi/revision_directive/index_en.htm
103
Market Assessment of Public Sector Information
The proposed European Data Protection Regulation
In January 2012 the European Commission published a draft General Data Protection Regulation. If adopted,
this would replace the current Data Protection Directive (95/46/EC) and any relevant member state
legislation, including the UK’s Data Protection Act. Unlike a Directive, a Regulation takes direct effect with no
need to be transposed into law for each member state.
There are aspects of the Regulation, in its initial draft form as proposed by the European Commission, that
have caused concern among holders and users of public sector information. In particular, the draft was seen
to favour the right to privacy of individuals and one possible outcome of this is that the Regulation may
impose restrictions and conditions on the linking of datasets and use of ‘big data’, where this involves
personal data.
In its proposed form the Regulation introduces new procedures for the processing of personal data. These
may increase the bureaucratic burden on organisations and could disincentivise the release of public sector
information where this becomes too burdensome. This is significant since the use of anonymised personal
data is among the most potentially valuable exploitations of public sector information, both for improving
policy and for wider welfare gains.
The Regulation appears to provide derogations from particular requirements for the use of personal data for
historical, statistical and scientific research purposes. To qualify for these derogations, personal data must be
processed in accordance with Article 83: personal data should be used only if anonymous data is insufficient,
and where possible identifying information should be kept separate from other information. This may make
the use of public sector information for research and policy purposes easier, although other areas of the
proposed Regulation may continue to pose challenges.
The Regulation is currently undergoing consideration and amendment by the European Parliament and
Council of Ministers, with adoption likely either during the course of 2013 or possibly in 2014.
The INSPIRE Directive
The Inspire Directive (2007/2/EC), which came into force in May 2007, is intended to enhance
availability, access and sharing of spatial information across EU member states. It is undergoing
implementation in stages, with full implementation required by 2019. The essential objective is to
enable sharing of geospatial information among public sector organisations, and improve public
access to this information. The information to which the regulation applies falls with 34 ‘themes’. 134
INSPIRE is based on the following principles:

data should be collected only once and kept where it can be maintained most effectively;

it should be possible to combine seamless spatial information from different sources across
Europe and share it with many users and applications;

it should be possible for information collected at one level/scale to be shared with all
levels/scales; detailed for thorough investigations, general for strategic purposes;

geographic information needed for good governance at all levels should be readily and
transparently available; and

easy to find what geographic information is available, how it can be used to meet a particular
need, and under which conditions it can be acquired and used. 135
134
Available to view at http://inspire.jrc.ec.europa.eu/index.cfm/pageid/2/list/7
135
See http://inspire.jrc.ec.europa.eu/index.cfm/pageid/48
104
Market Assessment of Public Sector Information
This is an example of data sharing best practice which could productively be replicated for other
types of public sector information. However, it also indicates the difficulties involved in achieving
such agreement. It took six years for the Regulation to be agreed upon; and full implementation
was not scheduled until a further 12 years after the date of agreement, in 2019.
Chapter summary

Despite the public sector information market being relatively young and still maturing, there are a number of
key policymakers and policy ‘shapers’ supporting and influencing its evolution.

The ability and inability of PSIHs to publish and share public sector information is influenced by express
statutory powers, implied statutory powers and common law powers.

The UK Government Licensing Framework provides the policy and legal overview of the arrangements
behind the use and re-use of public sector information and includes an Open Data-style licence.

The Information Fair Trader Scheme (IFTS) accreditation scheme, which is open to all public sector bodies,
has been designed to ensure that re-users of public sector information are treated fairly by PSIHs, and
contains principles around complaints handling, transparency and simplicity.

Exceptions to marginal cost pricing for public sector information can be made in certain cases where
specified criteria are fulfilled.

The Data Protection Act 1998 does not apply to truly anonymised or pseudonimised data and there are a
number of examples of how this can be done.

The key European regulations affecting public sector information in the UK are currently being revised with
final versions expected to be adopted later in 2013.
105
Market Assessment of Public Sector Information
5. The current value of the public
sector information market
There is no single recognised methodology for estimating the value of
public sector information and previous estimates vary considerably.
Further, the lack of data on the number of public sector information
datasets, consumers’ willingness to pay and usage statistics makes any
estimate subjective due to the need to make a number of modelling
assumptions.
Nonetheless, based on the available information and following a
number of informed, conservative assumptions developed with
stakeholders and on the basis of previous research, this report has
applied a pragmatic approach to estimate the value of public sector
information market in the financial year 2011/12. While it adapts a
previously used methodology, it makes a number of changes to reflect
the need to capture the entire public sector information market rather
than just an individual segment. This means that the estimates
presented here are not readily comparable with previous estimates of
the value of public sector information in the UK 136 .
The primary focus of the estimation exercise has been to calculate the
so-called narrow economic value of public sector information, which
captures the direct value to consumers and PSIHS as well as the
supply-chain and consumer spending value impacts. However, given
that public sector information is known to have significant broader,
social value impacts, case study analyses have been used to illustrate
this current value.
In some instances, these case studies are augmented with ad hoc
calculations that provide an indication as to what the broader societal
value of public sector information might be in financial terms.
136
For example, the figures estimated here are not directly comparable with the £16 billion figure quoted in the Open
Data White Paper.
106
Market Assessment of Public Sector Information
Understanding the value of public sector information
At the outset, it is important to clearly define what is understood by the term value 137 . Previous
studies have considered value in terms of market value (defined as either total turnover or profits
accruing to consumers of data and PSIHs), consumer 138 and producer 139 welfare and broader
(non-monetary) value. Having reviewed the evidence, it appears appropriate to disaggregate value
into four distinct components which can capture elements of market, consumer, producer and
wider value:

the direct value of public sector information to producers and suppliers (the PSIHs):
these are the benefits accruing to producers and suppliers of public sector information through
the sale of public sector information or related value-added services;

the indirect value of public sector information arising from its production and supply:
the benefits accruing up the supply chain to those organisations interacting with and supplying
PSIHs (but not directly using or re-using public sector information), and the benefits accruing to
those organisations where employees of PSIHs and supply chain organisations spend their
wages;

the direct use value of public sector information to consumers of public sector
information: the benefits accruing to businesses, civil society, individuals and the public sector
from directly using and re-using public sector information for a variety of purposes; and

the wider societal value arising from the use and re-use of public sector information: the
benefits to society of public sector information being exploited, which are not readily captured
elsewhere.
The first three types of value can be termed economic value or narrow economic value and can be
measured using standard economic methodologies to derive a monetary estimate for value and the
associated employment figure (see below).
The final type of value is harder to measure as it captures wider benefits arising from the use of
public sector information – these are typically not measured in monetary terms. The literature
discusses a number of ways in which public sector information can have broader impacts:

increasing democratic participation: giving citizens and businesses access to public sector
information allows them to perform their own analyses of salient issues, make more informed
choices about public service providers and interact with policymakers to challenge their
assumptions and improve the policymaking process;

promoting greater accountability: for example through the scrutiny of costs of public service
provision and benchmarking comparable services;

greater social cohesion: for example, by providing more information on the provision and
distribution of services, public sector information can be used to dispel myths on who receives
certain public services;

generating environmental benefits: such as reducing congestion and pollution through the
release of better traffic and transport data which helps drivers to better plan journeys; and
137
This is distinct from ‘value-add’ services that might be supplied by PSIHs using public sector information as an input.
This is the value or benefit consumers of public service information enjoy over and above the price they pay for it
(including if the price is zero).
139
This is the value accruing to PSIHS when public sector information is purchased by consumers. It is typically
analogous to profit received.
138
107
Market Assessment of Public Sector Information

identifying previously unknown links between different policy areas: through data-mash
ups it may be possible to develop system-wide solutions that holistically seek to address the
root of policy challenges.
Thus, for the purposes of this report, the total value of public sector information is defined as the
benefits accruing to the suppliers, users and re-users (i.e. consumers) of the information and data
in terms of profits generated, jobs created and supported (narrow economic value) and the wider
benefits to society arising from the exploitation of public sector information.
It is important to clarify that value defined in this way is not directly the same as the market value of
public sector information. Market value typically refers to the volume of sales multiplied by the price
of the product. In this way it includes labour costs, capital costs and the intermediate costs of
production which are not included in the Deloitte definition. Conversely, market value will not
capture indirect effects and wider societal benefits.
Modelling approach and assumptions
Appendix 5 contains full details of the modelling approach and assumptions used. This report
adopts a three-stage approach to valuation of public sector information for the financial year
2011/2:

Stage 1: estimating the value of public sector information to PSIHs and the value to direct
consumers (users and re-users) of public sector information using a bottom up approach that
quantifies consumer and producer surplus; and

Stage 2: estimating the value of the associated indirect and induced impacts to PSIHs using
Input-Output multipliers.

Stage 3: estimating a ‘ready-reckoner’ value of wider value based on other available research.
The report does not seek to systematically quantify the value of the broader social impacts of
public sector information. There is a lack of reliable evidence on the linkages (correlated and
causal) between the consumption of public sector information and democratic, social,
environmental and political impacts.
However, notwithstanding this caveat, a third stage of the approach includes a number of case
study analyses to illustrate the types of impacts that public sector information can have – and in
some cases these include ad hoc calculations to give an indication of the quantum of value that is
being generated. This is then used in conjunction with ‘uplift’ ratios taken in other studies to allow
us to generate a broad, order-of-magnitude estimate for wider, and thus the total value of public
sector information in the UK.
Again, it should be stressed that the lack of reliable data on public sector information 140 makes the
quantification of value difficult and heavily reliant on assumptions. To reflect this inherent
uncertainty the estimates contain upper and lower bounds. Further research (and especially data
collection) is required to develop more comprehensive estimates.
Stage 1: consumer and producer surplus approach – the welfare approach
The report uses the so-called bottom-up approach developed by the OFT/DotEcon (2006) 141 . As
the OFT/DotEcon report discusses, a bottom-up approach is preferable to a top-down approach
given the latter’s tendency to over-attribute causality and generate biased estimates. This ‘welfare’
approach seeks to quantify the consumer surplus 142 derived from the use and re-use of public
140
Namely, the total number of public sector information datasets held by PSIHs, the number of users by dataset and
how users are exploiting public sector information.
141
This report is publicly available and the methodology replicable once the necessary data has been sourced.
142
Defined as the value or benefit consumers of a product or service enjoy over and above the price they paid for it. It is
the difference between the price consumers pay and the maximum price they are willing to pay
108
Market Assessment of Public Sector Information
sector information and the producer surplus 143 accruing to PSIHs from the generation, collection
and dissemination of public sector information in the financial year 2011/12.
Following the DotEcon approach, the welfare approach can be estimated from summing the
current net surplus with the total producer surplus from the supply of public sector information.
There are standard formulae to calculate consumer surplus and producer surplus (see Appendix
5). The calculations use data on revenues attributable to public sector information, number of
known public sector information datasets and available download data. Assumptions are made
relating to:

price elasticity of demand for different dataset categories;

the proportion of users going on to actively use and re-use downloaded/viewed datasets; and

the shape of the demand curve.
The shape of the demand curve has an important bearing on the size of the consumer surplus.
While it may be the case that there are linear demand curves for public sector information that
carries a fee, the demand curves may exhibit non-linearities and discontinuities when the public
sector information is available free of charge – the level of consumer surplus may be much larger
or smaller when the demand curve is not linear.
This modelling exercise has sought to recognise these uncertainties and the upper and lower
bounds alter our assumptions across the three above variables.
Further details of the modelling approach taken can be found in Appendix 3.
Stage 2: indirect and induced value of public sector information
The indirect value of public sector information refers to the value generated in associated
industries that supply inputs into PSIHs. The induced value of public sector information refers to
the value generated when households spend wages paid out by PSIHs and industries involved in
the supply chain. The report estimates the size of these value impacts using the UK Domestic Use
Matrix for 2005 (latest available) sourced from the ONS and the direct use estimate derived above.
Appendix 5 contains more details of the approach, but broadly speaking it has involved taking the
direct use estimate of producer surplus from Stage 1 (analogous to operating profit) and converting
this into gross output (GO) and gross value added (GVA) on the basis of available data. GO TYPE
I and TYPE II multipliers are then used in an Input-Output setting to consider the upstream
business-to-business purchasing effects (indirect) and consumer spending effects (induced). The
GO multiplier used in this process was estimated to be 3.0.
To allow a comparable estimate of value to the original producer surplus (and thus allow
aggregation), the results are converted back into GVA and then operating profit, by definition,
providing a comparable surplus estimate. The ‘surplus multiplier’ – the ratio of indirect and induced
surplus to the direct use surplus – was estimated to be 1:2.4 (3.4 in total). In other words, for each
£1 generated as producer surplus in direct use, a further £2.40 is generated via indirect and
induced effects.
Per worker productivity estimates are used in conjunction with estimates of GVA to provide an
indicative level of employment supported in organisations supplying inputs to the public sector
information supply chain and supplying goods and services to consumers.
143
Defined as the value accruing to producers of goods and services when their output is purchased by consumes. In
traditional supply and demand analyses, producer surplus is calculated as the difference between the lowest amount the
producer would be willing to sell the good/service for and the price the producer actually sold it for. In many cases this is
equal to profit.
109
Market Assessment of Public Sector Information
Narrow economic value estimates
Taking account of the above caveats and limitations, the estimates of the current value of public
sector information in the UK in 2011/12 are shown below.
Figure 5.1: estimates of the current value of public sector information in the UK in
2011/12 (annual value)
Value category
Central scenario
High scenario
Low scenario
Direct consumer surplus
£1.6 billion
£2.00 billion
£1.00 billon
Producer surplus
£0.1 billion
£0.1 billion
£0.1 billion
Indirect and induced
value (supply chain and
consumer spending)
£0.1 billion
£0.1 billion
£0.1 billion
Total
£1.8 billion
£2.2 billion
£1.2 billion
Source: Deloitte analysis. Note these figures are not comparable with other estimates of the value of public sector information (such as
the previous DotEcon estimates) due to differences in the scope of analysis and methodology used.
Note the calculated values for producer surplus and indirect and induced impacts do not alter
between scenarios as the data is more robust and fewer assumptions are made. Changes to
elasticity, demand curves and data only affect direct consumer surplus estimates.
As noted in the introduction, these figures are not comparable with previous estimates of the value
of public sector information as earlier attempts have focused on specific datasets or sought to
include selected social impacts.
As the review of evidence has shown, public sector information can generate value across a much
wider range of indicators – suggesting that the total current value of public sector information far
exceeds £1.8 billion. The following section considers the wider social value of public sector
information in the UK.
Wider societal value estimates
The use and re-use of public sector information can generate a range of benefits, which in turn can
create value for consumers and producers alike. Some of the types of benefits that have
demonstrable and, in some cases, calculable financial implications include:

the value of reduced carbon emissions, reduced fuel use and time saved due to reduced
congestion through using apps and other tools that rely on live public sector information APIs,
for example to give live transport updates;

the value of efficiency savings in the public sector 144 through identifying cost savings arising
from analysis of different public sector information datasets;

improvements to decision making and choice due to new insights from public sector
information which, in turn, helps reduce barriers to entry for new entrants to different markets,
helping generate economic growth;
144
It is assumed, based on the evidence reviewed, that cost savings arising from the use and re-use of public sector
information are likely to exceed value, as defined. Consider a hypothetical saving of £1 billion through using public sector
information to reduce healthcare costs. This benefits the Exchequer and gives the Government an additional £1 billion to
spend or an equivalent amount in tax cuts. However, what it can also do is reduce value as defined by producer surplus
as healthcare companies may see face falls in profit and they may choose to reduce R&D. The exact balance will
depend on international flows of value and how the £1 billion is redistributed.
While the evidence suggests the cost savings will be greater than any fall in value, it is for these reasons of uncertainty
that this impact has been considered separately from other narrow economic impacts.
110
Market Assessment of Public Sector Information

promoting greater accountability in public services and public life through the availability of
public sector information datasets that help individuals rate performance and pursue
accountability; and

better policy making using public sector information datasets that improve value for money
and the efficacy of policy.
To illustrate the ways in which public sector information currently generates value, the report
presents a number of case studies in industries already exhibiting some of the benefits of using
and re-using public sector information:

the healthcare and life sciences sectors;

the transport sector 145 ; and

the public sector.
Where appropriate, indicative estimates of the financial value being generated by public sector
information are provided. It is not, of course, always possible to measure benefits in financial
terms. Figure 5.2 on the following page offers an indication of the types of benefits and impacts.
Note that this includes only the benefits from the discussed case studies, and should be treated as
indicative of the much larger total value of the wider societal value of public sector information. A
summary of the case studies is also included on the following pages; see the appendices for
further detail on the sources and any empirical methodologies used for these.
Whilst these individual estimates of social value from the different case studies cannot simply be
summed (some are cost-savings, some are time-savings some are economic value and some are
non-quantifiable), they suggest that the total social value of public sector information is likely
to be significantly greater than the narrow economic value presented here. A conservative
estimate of the wider social value of public sector information suggests a value at least three times
the size of the £1.8 billion economic value may be appropriate 146 , i.e. over £5 billion p.a.
145
A ‘deep dive’ in this sector is placed in Appendix 6.
This is based on insights from individual case studies by other authors which have variously estimated the wider value
from public sector information use in the health care sector, geo-spatial, meteorological and other sectors in the order of
billions.
146
111
Figure 5.2: wider societal benefits of public sector information
Source: Deloitte analysis
This chapter concludes with some ‘blue skies’ opportunities for the use and re-use of public sector
information, which are not currently producing tangible benefits but which studies indicate may
prove valuable opportunities in the future.
The health sector
As is well-known, the health sector collects a wealth of data on patients and patient outcomes. The
case studies below highlight some of the opportunities that can arise out of analysing public sector
information datasets, but also some of the challenges around making this data more widely
available.
Publication of mortality rates following cardiac surgery
Case study summary:
Publishing data on mortality rates following adult cardiac surgery appears to be associated with a decline in
mortality. There are various theories as to why this is the case, including competitiveness among surgeons
which leads to a rise in performance, increased awareness among healthcare professionals, and public
pressure for higher standards. However, there are also concerns that the apparent decline in mortality may
reflect ‘gaming’ of the mortality data.
Size of the prize:
A decline in mortality rates is, in itself, desirable. The economic value depends on the value attributed to a
statistical life and the costs of providing treatment under various outcomes/pathways. Taking a median value
from a range of recent studies, the value of lives saved among those undergoing adult coronary artery surgeries
in NHS centres in north-west England in 2005 was around £55 million. This suggests that the total value of lives
saved per annum could exceed £400 million for England and Wales, if similar benefits were observed in all
regions.
Mastodon C – identifying NHS prescription savings from big data
Case study summary
By using data on prescribing practice across England, variations in spending on different classes of drugs can
be identified. It is then possible to calculate the potential savings to be achieved by moving from prescribing
branded to generic drugs.
Size of the prize
For statins alone, the study found that the NHS could save around £200 million per year by reducing
prescriptions of branded in favour of generic versions.
When extended to all classes of drugs, the total potential savings could amount to £1.4 billion per year,
according to work by the British Medical Journal.
A patient database for the NHS – the challenges to extracting value from large datasets
Case study summary:
A central NHS patient database could offer significant savings to the NHS as well as improving standards of
care and the patient experience. However, there are significant technical, privacy and cultural hurdles to
overcome if this is to be made a reality. This case illustrates both how attractive the prize of harnessing the
power of large public sector information datasets can be, but also the difficulties these can present.
Size of the prize:
Difficult to estimate, but if the system is delivered as planned the savings are likely to run to billions of pounds. An
initial illustrative estimate identified £4.4 billion of potential savings, although these are not all directly related to
the patient database.
The transport sector
The transport sector is an example of rich data with a high velocity. The case studies below show
how this data is being used to generate both narrow economic value and wider societal value.
Market Assessment of Public Sector Information
The break-out box below illustrates some of the areas of research and policy where sharing of data
could generate new insights and lead to improved policy outcomes. 147
Transport for London data
Case study summary:
Transport for London has released considerable volumes of information on the London transport network,
including live network updates. This case study assesses the value of time saved due to avoidance of
disruption on the transport network by travellers, using data published by Transport for London and accessed
via mobile apps.
Size of the prize:
The value of time saved due to avoidance of disruption in 2012 is estimated at between £15 million and £58
million, depending on the modelling assumptions used. This is likely to be a conservative estimate: for
example, the figure would be higher if a higher value for working time were used instead of a valuation based
on commuting. It also omits other aspects of value generated by Transport for London data, for example in
routine (non-disrupted) journeys by allowing travellers to plan their journeys more efficiently.
Nonetheless, this order of magnitude represents a significant annual time saving. By way of comparison, the
HS2 impact assessment calculates operational time savings equivalent to £440 million per year in 2012 values.
Therefore releasing relatively low-cost data, which is in many cases a by-product of other activities, may
generate time savings versus a nominal baseline valued in excess of 10 per cent of the time savings resulting
from a major national infrastructure project such as HS2.
Traffic England
Case study summary:
Traffic England offers live traffic data to road users, enabling users to plan their journeys so as to avoid
congestion, roadworks and other conditions likely to cause delays. They are thereby able to save time and
reduce fuel waste.
Size of the prize:
As a high level estimate, if 50 per cent of monthly users save ten minutes each on their journeys over the
course of a month, the value of time saved could equate to £6.5 million per year. This is likely to increase as
more road users become aware of the service, and as increasing congestion on the UK road network increases
the value of live traffic information.
Improving access to fragmented information – roadworks.org
Case study summary:
By providing a platform that combines information on planned and on-going roadworks into a single database,
roadworks.org provides a resource for road users that allow them to easily see potential disruptions to their
journey. It also reduces costs to local authorities from their own bespoke platforms, and improves
communications between utilities companies and local authorities.
Size of the prize:
A recent ELGIN report estimated the total benefits at around £25 million per annum. This includes ‘tangible
savings’ of £6.3 million to local authorities as a result of efficiency savings, and ‘intangible benefits’ of £19
million due to reduced congestion.
The public sector
The final set of case studies examine how the public sector itself is beginning to exploit public
sector information to drive efficiencies, improve public services and enhance policymaking.
147
The UK Administrative Data Research Network: Improving Access for Research and Policy (December 2012)
114
Market Assessment of Public Sector Information
Area of research and
policy
Description
Social mobility
Linking data on education, training, employment, unemployment, income and
benefits
Causal pathways over the
life course
Linking data on education, health, employment, income and wealth
Support for the elderly
Comparative analysis of access to and provision of social care support for the
elderly
Poverty
Linking data on housing conditions, health, incomes and benefits
Social care for children
Linking indicators of parental employment, social background and childcare
Offence and re-offence
Linking data on offending and re-offending behaviour, income, benefits, health
and mental health
An example of data sharing at a local level – families with complex needs
Examples of data sharing between local councils and other public bodies are proliferating across the UK. These
are often responses to the twin pressures of deep funding cuts and intractable problems involving multiple
agencies.
An example of this sort of problem is the case of families with complex needs, often referred to in the press as
‘troubled families’. These families may combine issues such as mental health problems, children out of school,
and long term worklessness and benefit dependency, meaning that they fall within the remit of multiple public
sector bodies including social services, the Police, and the local and national welfare services. These services
are estimated to cost around £75,000 per family per year. 148
Increased coordination between local councils and other local partners may prove both more cost efficient and
more effective in resolving the problems faced by such families. The councils of Greater Manchester,
Leicestershire and Bradford are working together to improve information sharing and management in this area.
The project aims to develop a single toolkit for information sharing, combining existing guidance and
approaches. This is intended to be applicable to both the councils and the agencies working with families with
complex needs.
In addition the project is intended to lay the foundations for a culture more conducive to information sharing,
addressing issues such as different professional cultures of sharing, lack of training and expertise, and differing
interpretations of legislation.
In practice, the steps needed to achieve this can appear prosaic but are nonetheless potentially powerful
enablers of an environment in which information can more easily be shared within and between organisations.
As the project is still ongoing it is too early to judge its effects, whether in terms of reduced costs or improved
outcomes. Nonetheless, this appears to be a positive example of the potential of greater sharing and
exploitation of data between councils and other public sector bodies in response to a policy problem which has
proved unresponsive to a siloed approach.
148
Association of Greater Manchester Authorities, ‘Improving Information Sharing and Management: A National
Exemplar Project’
115
Market Assessment of Public Sector Information
The Justice Data Lab
It is currently difficult for many providers of offender services, particularly in the voluntary and community
sector, to access re-offending data relevant to the offenders they work with. As a consequence, organisations
may encounter significant difficulties in measuring the effectiveness of their rehabilitation work with respect to a
reduction in re-offending. The lack of access to high-quality re-offending information has also prevented some
organisations learning from and improving the services they deliver; and has made it difficult or impossible for
them to demonstrate their impact to commissioners.
In order to address this problem the MoJ has set up the Justice Data Lab, announced by Secretary of State,
the Right Honourable Chris Grayling MP in December 2012. It is currently in a pilot phase running for a year
from April 2013. The Justice Data Lab will provide organisations with aggregate re-offending data specific to
the offenders they have been working with, and that of a matched control group. This is intended to allow them
to understand their specific impact in reducing re-offending. The intention is that supporting organisations by
providing easy access to high-quality re-offending information will allow them to focus only on what works,
better demonstrate their effectiveness, and, it is hoped, ultimately cut crime in their area.
Participating organisations will supply the Justice Data Lab with details of the offenders they have worked with
and information about the services they have provided. The Justice Data Lab will supply aggregate one-year
proven re-offending rates for that group, and a matched control group of similar offenders. The re-offending
rates for the organisation’s group and the matched control group will be compared using statistical testing to
assess the impact of the organisation’s work on reducing re-offending. The results will then be returned to the
organisation with explanations of the key metrics, and any caveats and limitations necessary for interpretation
of the results.
This is an example of how data of a personal, sensitive nature may be used to inform those who have a
specific need for this data. It indicates that these barriers can be overcome in a manner that does not
compromise ethical considerations around data protection and the right to privacy, while still allowing the data
to be used to improve policy and service delivery including by third parties. It also demonstrates the notion of
data being a ‘two-way street’ between PSIH and user, noted elsewhere in the study, which can lead to
improved outcomes for both parties.
However, this example also indicates that some investment into data analysis capability is likely to be required
on the part of the public sector information holder (in this case the MoJ) – in other words, there is a cost
involved. Depending on the success of this pilot over the course of 2013-14, it may be that this provides a
model for granting controlled access to similar datasets held across the public sector.
Source: www.justice.gov.uk/downloads/justice-data-lab/justice-data-lab-user-journey.pdf
‘Blue sky’ opportunities
In addition to the specific examples already discussed, there are many areas in which public sector
information may have a positive impact in the future. Foresight, part of the Government Office for
Science, 149 has carried out a number of studies which highlight the role public sector information
could play in improving our understanding of and ability to respond to the challenges our society
confronts. These are summarised below. 150
Detection and Identification of Infectious Diseases (2006)
Novel Information Technology for the Early Detection of Infectious Disease events. This project
identified a ‘Grand Challenge’ around using modern information and communication technology
systems to gather and interpret timely and relevant data, and to deliver it to those managing an
outbreak.
Tackling Obesity (2007)
On-going data collection and evaluation relating to the effect of different interventions was
proposed as a core principle in tackling obesity. This was seen as a particularly important in view
of the uncertainties surrounding the efficacy of certain interventions. The project stimulated the
149
150
See www.bis.gov.uk/foresight
With thanks to the Government Office for Science
116
Market Assessment of Public Sector Information
creation of the National Obesity Observatory - but much more remains to be done. Mismatches
between policy and research timescales remains a fundamental challenge.
Global Food and Farming (2011)
Policy makers are hampered by conflicting results from different models associated with the food
system. The project argued the need to substantially improve datasets that provide the basis for
model calibration and comparison. The project brought together a forum of modellers to help take
this forward.
Computer Based Financial Trading (2012)
Improving collection and access of financial transaction data was a key recommendation: a trusted
‘European Data Centre’ was proposed. Such data would be a crucial new tool for regulators in
identifying abuse spanning different markets. Also, if made easily available, it could unlock the
resources of the scientific community to better understand evolving markets, and so help policy
makers in the ‘arms race’ with algorithmic trading developers.
Reducing the Risks of Future Disasters (2012)
The report argues the need to improve the infrastructure for data collection for hazards: e.g. better
coordinated effort on satellites and sensors. It also proposes establishing a much needed database
of evidence on the costs and benefits of interventions to reduce disaster risk. Importantly, the data
in such a repository would need to be quality assured, and easily accessible to decision makers at
different levels in different countries.
The Future of Identity (2013)
The report considered how notions of personal and social identity in the UK might change over the
next ten years. Policies will need to take into account the multiple nature of identities and how
policies might affect groups differently, or individuals’ different times and places. The report
explicitly considers the commercial value of identity through the use of ‘Big Data’. This is likely to
become crucial to private sector organisations, but also has the potential for criminal exploitation,
for example through opportunities for identity theft.
117
Market Assessment of Public Sector Information
Chapter summary

The value of public sector information can be disaggregated into four broad categories according to the
beneficiary:
o
the direct value of public sector information to producers and suppliers (the PSIHs): these are the
benefits accruing to producers and suppliers of public sector information through the sale of public
sector information or related value-added services;
o
the indirect value of public sector information arising from its production and supply: the benefits
accruing up the supply chain to those organisations interacting with and supplying PSIHs (but not
directly using or re-using public sector information), and the benefits accruing to those
organisations where employees of PSIHs and supply chain organisations spend their wages;
o
the direct use value of public sector information to consumers of public sector information: the
benefits accruing to businesses, civil society, individuals and the public sector from directly using
and re-using public sector information for a variety of purposes; and
o
the wider societal value arising from the use and re-use of public sector information: the benefits to
society of public sector information being exploited, which are not readily captured elsewhere.

The first three value categories can be grouped together as narrow economic value and have been
estimated at £1.8 billion p.a. for the year 2011/12.

The wider societal value of public sector information is harder to estimate. Public sector information can
benefit society through a number of routes including:
o
increasing democratic participation: giving citizens and businesses access to public sector
information allows them to perform their own analyses of salient issues, make more informed
choices about public service providers and interact with policymakers to challenge their
assumptions and improve the policymaking process;
o
promoting greater accountability: for example through the scrutiny of costs of public service
provision and benchmarking comparable services;
o
greater social cohesion: for example, by providing more information on the provision and
distribution of services, public sector information can be used to dispel myths on who receives
certain public services;
o
generating environmental benefits: such as reducing congestion and pollution through the release
of better traffic and transport data which helps drivers to better plan journeys; and
o
identifying previously unknown links between different policy areas: through data mash-ups it may
be possible to develop system-wide solutions that holistically seek to address the root of policy
challenges.

This value is potentially significant and is likely to have a major influence on overall societal wellbeing. A
conservative estimate of the wider social value of public sector information suggests a value at least three
times the size of the £1.8 billion economic value may be appropriate, i.e. over £5 billion p.a.

Adding this wider social value estimate to the calculated value of public sector information to consumers,
118
Market Assessment of Public Sector Information
businesses and the public sector, gives an aggregate estimate of between £6.2 billion and £7.2 billion in
2011/12 (2011 prices).
119
6. Barriers and policy implications
Using a taxonomy originally developed by the Data Strategy Board, this
chapter considers the current barriers to the UK in fully realising the
benefits and value of public sector information. The analysis is based
on discussions with stakeholders and a review of the evidence.
The chapter also considers the policy implications of each barrier. The
Shakespeare Review will be making its own recommendations on which
barriers and associated challenges it believes are the most important
and the steps that can be taken to address these.
Generating value
Having reviewed the literature and evidence, it is apparent that the two main ways in which value is
being generated currently, which will likely remain the case in the future, is through data discovery
and data exploitation. The former relates to making better and, in some cases more, public sector
information available and making it more accessible – what might be loosely term supply-side
considerations. The second dimension relates to using and re-using public sector information
better – demand-side considerations. This might be through reducing consumer risk aversion to
using public sector information, improving data exploitation techniques, changing cultures and
improving data analysis skills (which, in turn can improve competiveness).
Figure 6.1: exploiting the value of public sector information
Source: Deloitte analysis
The light blue boxes (which are not to scale) represent the value that is currently being generated
from public sector information, with the first light blue box on the left showing the current value from
data discovery and the second light blue box the current value through data exploitation and data
science. The dark blue boxes illustrate the additional value that could be achieved if (i) better
quality public sector information was released, (ii) more was made more easily available; and (iii)
more public sector information datasets were released. While, in some cases, more public sector
information being available can also increase value, it is important to note that there is not a linear
relationship between quantity of public sector information and its value. What is key is the quality of
Market Assessment of Public Sector Information
this data and its amenability to data analytics. Simply releasing more datasets, irrespective of their
quality, is likely to only have a minimal value impact.
More effective ways of data exploitation can include:

greater system-wide analysis through more sharing and combining datasets together to
consider policy and business issues from a range of non-traditional perspectives;

unlocking the potential from linked data: and better integrating; and

greater data-mashing with private sector and individual datasets.
Value is thus generated through exploring existing datasets to identify insights through statistical
analysis, ‘data-mashing’ and visualisations. The value of public sector information in the UK can
therefore be increased both by increasing the quality, and in some cases quantity, of datasets
available/accessible (increasing the potential for data discovery) or by better using existing and
new datasets (increasing data exploitation).
This raises the question of what the barriers to the UK fully exploiting the value of public sector
information are, and what the size of this potential value may be.
A taxonomy of barriers
This report adopts a taxonomy originally developed by the Data Strategy Board, which illustrates
the four main areas where challenges or barriers may currently exist. This taxonomy of barriers is
shown in Figure 6.2 below.
Figure 6.2: taxonomy of barriers
Source: adapted from DSB
Legislative barriers include whether certain acts of legislation and other regulations reduce the
usability of public sector information datasets by consumers. There are also further questions over
whether current assurance/accreditation standards for public sector information remain effective
and whether there is scope for existing licensing arrangements to be improved.
Economic barriers include questions as to which datasets to release (which datasets yield the
highest value) and how their costs (if any) should be covered.
Access barriers to maximising the value from public sector information can include:
121
Market Assessment of Public Sector Information

a reluctance to publish or share datasets;

a bias against using non-traditional datasets;

a reluctance to use public sector datasets because of concerns around quality, reliability and
on-going support; and

a lack of skills to fully exploit the value of public sector information.
It should be noted that the extent of these barriers may differ between public sector information
datasets and also between the different types of users.
Each of these barriers is discussed in more detail below, alongside the relevant policy implications.
Note, due to the overlap in some of the taxonomy areas, the narrative below does not always
follow Figure 6.2 exactly.
Legislative barriers
Privacy and the impact of current regulations and legislation
The review of available evidence suggests that, by and large, the current legislative and regulatory
environment around public sector information is not acting as a barrier to generating value and
market development. While a full legal analysis has been beyond the scope of this study, particular
acts and regulations such as those covering Data Protection and Human Rights legislation have
not emerged as preventing the development of new products or services using public sector
information 151 . The majority of stakeholders consulted have not reported that these generic
legislative acts are currently preventing them from using and re-using public sector information,
but, in some cases specific legislation controlling a particular data collection exercise do.
However, the evidence received does suggest there are some current challenges around
perceptions and attitudes towards data release. In particular:

regulations such as Data Protection are sometimes used as a shorthand justification for not
sharing public sector information within the public sector, with PSIHs not always able to
translate their awareness of their rights and duties into scenarios where public sector
information is released or shared, causing a barrier; and

when a policy decision is taken not to release public sector information datasets to the general
public, the reasons are often not well articulated or the conditions attached to access are overly
restrictive.
Some stakeholders have cited examples where the Data Protection Act (and other legislation) has
been used as a reason for a PSIH to withhold information from the general public. However, as
noted in previous chapters, in reality, the Act should rarely be a barrier to sharing information.
Where public sector information datasets do contain personal details, these may need to be
aggregated or anonymised. Where this is done effectively the Data Protection Act, in most cases,
no longer applies to the information and it can be made available.
This is not to downplay data protection issues. The impact of breaches leading to release of
personal information can be extremely serious. Further, even if data is anonymised, PSIHs may be
reluctant to release it if they believe it can have wider consequences, e.g. reporting crime data may
adversely affect house prices 152 . These issues notwithstanding, PSIHs are beginning to use
innovative methods to test different ways of anonymising datasets – for example the Ministry of
151
This observation should be caveated with the note that, at the time of writing, there are a number of initiatives to
revise various regulatory frameworks at the European level.
152
See for example: www.guardian.co.uk/money/2011/feb/01/police-crime-website-house-prices
122
Market Assessment of Public Sector Information
Justice worked with statisticians, the private sector and the academic community to avoid the
‘jigsaw effect’ 153 occurring from the release of offender data (see Chapter 5).
Access to the general public
With respect to other access restrictions to public sector information datasets (such as the datasets
only being available to researchers or in secure environments) arising from specific pieces of
legislation, in many cases the rationale for these restrictions are clear. The rationale may cover
national security reasons or genuine data protection concerns. In many cases this data is released
only to licensed researchers, and some stakeholders questioned the value of making it available to
the wider public since its applicability is to a specialised area of research. However, other
stakeholders have reported that in some cases, the rationale for restrictive access to certain
datasets is not always clear, overly restrictive or no longer relevant.
For example, some start-up companies have reported difficulties in accessing different versions of
the National Pupil Database as non-research organisations (see case study below). While
restrictions on access are in this case clearly needed to protect sensitive and personally
identifiable information; there is a valid question as to whether there are ways that some or all of
this information could be made available to commercial organisations to develop new products and
services to help parents, teachers and other stakeholders to improve decision-making, increase
accountability and identify good practice. Without this, opportunities for innovations using this and
other datasets are likely to be missed, meaning loss of the value that they could potentially create.
Case study: the National Pupil Database
The National Pupil Database (NPD) was created under the School Standards and Frameworks Act 1998,
which instituted a legal gateway permitting the collection of personal data about students without their
consent or the consent of their guardians. The information was to be drawn directly from school management
systems. Until the late 2000s students and parents were not informed that this information was being
collected and stored.
Since 1998, the National Pupil Database has expanded to contain ever more information about students.
From initially being an annual census, information is now collected every term, and has expanded to include
preschools, any provision of childcare that is funded by the state, exclusions of students and the reasons for
exclusion, recipients of free school meals, and the mode of travel to school., among other fields. Because the
initial gateway offered something approaching carte blanche in terms of the information that could be
collected, it has been gradually expanded according to the needs of policymakers.
This means that the NPD is tremendously rich as a source of data. It also means that it contains a wide range
of potentially sensitive and personally identifiable information. This information is stored permanently, often
without the knowledge and in any case without the consent of the subject.
The NPD therefore creates a dilemma. On the one hand, the insights that could be derived from this
information, especially when combined with other sources of information, are tantalising. For example, it
offers researchers and policymakers a route into understanding the drivers of educational attainment,
behavioural problems among children, and social mobility in later life. This information could also be used to
generate a range of value-added services and products by private companies.
On the other hand, there are clear risks associated with the holding and sharing of such a quantity of
sensitive personal information. For example, information about exclusions of pupils, including the reason, is
stored indefinitely. If this information were to be released in an identifiable manner it could have a
compromising effect on the individual in later life, for example when applying for work.
The Department for Education currently has a well-developed process in place for access to the information
to safeguard against negative outcomes. Those who require access are required to demonstrate that this
information is necessary research into the education achievements of pupils. The requests are processed by
the DfE Data and Statistics Division (DSD), with requests for Tier 1 data (information that is both identifiable
and highly sensitive) referred to the DfE Data Management and Advisory Panel. Where access is approved,
153
The process of combining anonymised data with auxiliary data in order to reconstruct identifiers linking data to the
individual it relates to.
123
Market Assessment of Public Sector Information
Case study: the National Pupil Database
the DSD manages the supply of data.
In November 2012 the DfE launched a consultation on widening access to the NPD, with a view to granting
access to: “persons (i) conducting research or (ii) providing information, advice and guidance or (iii) data
based products and services for the purpose of promoting the education or well-being of children in England
and who require individual pupil information for that purpose.”
However, these changes may still preclude some potentially valuable uses of the NPD. Restricting access
along the lines children’s ‘well-being’ potentially forecloses innovative non-educational analyses. For
example, a researcher exploring links between educational performance and culture or the environment
would have to seek to justify their use of the NPD in ‘well-being’ terms – this may not be straightforward.
Equally, always having to specify a reason for requesting the NPD may prevent speculative research or
‘randomised’ data-mash-ups.
Despite this, the personal nature of the information inevitably makes this a sensitive topic, as it may be felt
that widening access increases the risk of accidental disclosure of the information, as well as allowing
information to be used in ways the data subjects have not consented to. In its response to the consultation
the ODI cited a number of uses to which aggregations of data from the NPD could be used as inputs to
products and services developed by start-up business, but stated that “the aggregations these applications
require could be generated by the Department for Education. None of these services require access to
individual-level data.” 154
This case therefore illustrates the complexity of the barriers to access in cases of sensitive, identifiable
information, and the tension that may arise between the right to personal privacy and the potentially rich uses
of datasets containing personal information. At the time of writing, the DFE NPD consultation had closed with
a response expected soon.
Source: partially based on an ODI talk, ‘Getting to grips with the National Pupil Database, 15th February 2013.
Slides available at www.scribd.com/doc/125638490/Getting-to-grips-with-the-National-Pupil-Databasepersonal-data-in-an-open-data-world
See also www.education.gov.uk/researchandstatistics/national-pupil-database/b00212283/national-pupildatabase
The case study below illustrates the case where the value of public sector information remains
locked not for reasons of restricted access or cost, but simply because a given dataset is not being
made available.
Honest Buildings – ‘LinkedIn for the real estate market’
Honest Buildings is a start-up that originated in the United States but is now expanding into the UK, and to this
end is currently working with the ODI in London. It provides an example of how greater availability of public
sector information can be combined with private sector and crowd-sourced information to generate new
efficiencies in an industry sector.
The real estate sector – covering owners and occupiers of buildings, as well as providers of building services
and improvements – currently depends on a relatively cumbersome approach for most interactions between
parties. For example, businesses searching for new premises, or for improvements to their property are forced
to enter into a potentially lengthy and resource-intensive process of inviting potential suppliers to tender for
work.
Honest Buildings offers a platform similar to LinkedIn for the real estate industry, providing profiles for buildings,
organisations, projects and people. The idea is to build up a transparent, searchable network of information on
buildings, their owners and occupiers, suppliers of services and their project histories. This should make the
process of interactions between these parties much more efficient, as potential buyers of services will be able to
receive more relevant, informative tenders and suppliers will be able to adopt a more focussed approach to
business development. Michael Adler, former CFO and executive vice president of the travel website
Expedia.com, has said: “Honest Buildings is bringing the same type of efficiencies to the real estate market that
154
See www.theodi.org/consultation-response/proposed-amendments-individual-pupil-information-prescribed-persons
124
Market Assessment of Public Sector Information
Honest Buildings – ‘LinkedIn for the real estate market’
Expedia and Trip Advisor did to the travel industry.” 155
Much of the information on this platform will be crowd-sourced from building owners, occupiers and service
providers, but Honest Buildings also supplements this with public sector information in order to add additional
value. This includes property prices, rates, details of building specifications and energy efficiency details.
Honest Buildings
Source: http://www.honestbuildings.com/
Honest Buildings already have a presence in nine cities in the United States. In addition to their wider social
networking platform they have also found success providing platforms for energy efficiency programmes in New
York State and Connecticut. They are hoping to get involved with analogous initiatives in the UK like the Green
Deal, GLA’s RE:FIT programme, Cambridge Retrofit along with other major public/private sector led building
initiatives. The Green Deal allows consumers to have a range of energy savings improvements made to their
homes and then pay for the work through additions to their energy bill. This would allow the government and
suppliers to more efficiently target those buildings with an Energy Performance Certificate (EPC) of F or G, the
lowest ratings and therefore the highest priority for energy saving upgrades. RE:FIT and Cambridge Retrofit
provide a framework focused for non-domestic buildings to be upgraded in the public sector in London and in
the public and private sector in Cambridge respectively.
The potential economic benefits of widespread uptake of the Green Deal are significant. For example, this could
generate large savings on domestic energy bills. The Energy Saving Trust calculate that loft insulation of
270mm could save the average three bedroom house up to £180 per year on their energy bill, and double
glazing could save around £170 per year. 156 If 20 per cent of the UK’s approximately 25 million households
were able to achieve these combined savings of £350 per year, this would equate to an annual saving on
domestic energy bills of £1.75bn.
In addition, and more directly relevant to the current focus of Honest Buildings, there is also the potential for
savings on the energy bills of commercial properties. The scale of the potential savings is more difficult to
155
See www.greenbiz.com/blog/2012/09/12/secret-life-buildings
156
See www.energysavingtrust.org.uk/Energy-Saving-Trust/Our-calculations
157
See www.gov.uk/government/uploads/system/uploads/attachment_data/file/47986/1002-energy-bill-2011-ia-greendeal.pdf
158
Ernst & Young, ‘Making energy efficiency your business’, 2011,
www.ey.com/Publication/vwLUAssets/Making_energy_efficiency_your_business__Understanding_the_potential_of_the_nondomestic_Green_Deal/$FILE/EY_Making_energy_efficiency_your_business_-_Non_domestic_Green_Deal.pdf
125
Market Assessment of Public Sector Information
Honest Buildings – ‘LinkedIn for the real estate market’
estimate as there exists no comprehensive register of the UK’s commercial property stock. In addition
commercial buildings are more heterogeneous in nature, meaning that it is more complex to assess the benefits
of energy saving measures. As an indication, the 2010 DECC impact assessment for the Green Deal calculated
potential energy savings of between £170 million and £330 million, based on an additional uptake of energy
saving measures of 10 to 20 per cent above a business as usual scenario. However, these savings are set
against capital costs of between £75 million and £140 million. 157
These are indicative estimates, but demonstrate the scale of the potential saving from widespread uptake of the
Green Deal. They also omit the downstream impact, including improved energy security due to less reliance of
imported energy, reduced carbon dioxide emissions and other negative environmental externalities, and greater
consumer spending power due to reduced energy bills.
In addition, there is likely to be a wider economic uplift due to the generation of green jobs, support for the
construction industry, and increased demand for raw materials, many of which can be sourced within the UK
and would therefore not need to be imported. A recent study on the non-domestic Green Deal indicated a
potential market size, for SME uptake of Green deal measures, of between £470 million and £800 million by
2020, depending on the rate of uptake. 158 This would provide a substantial economic stimulus to this sector of
the economy.
There is, in addition, a potential legislative imperative to upgrading building energy efficiency. It is proposed to
ban the rental of buildings gaining an F or G EPC rating. According to Honest Buildings this applies to an
estimated 20 per cent of the UK’s building stock, and at the current time such a ban would therefore throw the
real estate market into crisis. Upgrading this stock is for this reason likely to become a priority over the next
decade, and a task of this magnitude will require an efficient and transparent approach, with property owners
fully aware of their supplier options and the relative costs and benefits of each.
Honest Buildings therefore appear to be an example of innovation using public sector information which could
generate a range of tangible economic benefits, from energy savings and job creation to reducing transaction
costs across an entire sector. However, there are barriers to accessing the information required. Most
importantly, EPC register, administered by Landmark Information Group, is not currently accessible. Honest
Buildings report that Landmark claim they are unable to give access to the database due to their contractual
obligations to CLG.
The EPC register is an example of ‘by product’ data – although there might be a small cost associated with
making it publically available (due to hosting requirements and user support), there would be no additional
collection costs as this is data that is collected in the course of the normal operations of the EPC scheme.
Release of this data therefore appears to be a win-win option which could unlock considerable potential value
without leading to a loss of revenue or incurring significant costs to the information holder.
More generally, Honest Buildings has identified issues with fragmentation and disconnect across the public
sector’s approach to information. It has frequently found it difficult to find the information needed, and that
where this has been possible to locate, access has sometimes been stymied by confusion over permissions
and ownership, meaning that information holders do not feel empowered to release the information. This
reinforces the argument for greater coordination and clarity in the public sector’s approach to information, with
an ‘open by default’ assumption for information and a clear framework defining the rules around information
release. Support for providing information about what data is available, whether through centralised catalogues
or decentralised search, would also help guide potential users to the information they need.
In addition to the barriers discussed above, there has been much discussion about core reference
datasets in the context of the debates about releasing the Postcode Address File 159 .
Access within the public sector
Within the public sector, there are also questions surrounding the legal ability of different public
bodies to share datasets with one another. These have been explored in depth in the recent report
by the Administrative Data Taskforce 160 . It notes that as regards the legal gateways established
to allow departments and other public sector bodies to share information without obtaining the
consent of the data subjects, “recent experience demonstrates that link-specific gateway legislation
159
160
See for example: http://data.gov.uk/blog/odug-progress-on-a-national-address-dataset
Source: Administrative Data Taskforce (2012)
126
Market Assessment of Public Sector Information
is both cumbersome and inefficient.” As a solution to this the Taskforce recommends the creation
of a generic legal gateway to reduce this barrier to the exploitation of public sector information and
clarify the legal position around sharing data. Note that this applies to sharing of data within the
public sector and with certain accredited third parties such as researchers, rather than publication
for use by the general public.
Insights
 Existing regulations and legislation governing public sector information do not appear
to be acting as actual barriers to realising the full potential value of public sector
information. However, the manner in which some of these regulations and legislation
are interpreted can lead to overly risk averse behaviour and can create barriers.
 Increasingly effective anonymisation techniques and an approach across the public
sector that emphasises granting access to as much data that is compatible with
privacy and security has the potential to improve access to public sector information.
 The eligibility restrictions imposed around certain datasets are typically due to
reasons of national security, data protection and other sensitivities.
 The rationale behind these restrictions is not always clearly articulated and may, in
some cases, no longer apply. In some cases, this may be preventing opportunities for
innovation to take place.
Licences
Licensing conditions play an important role in facilitating (or preventing) the full exploitation of the
value of public sector information. The ideal standard widely acknowledged by stakeholders is
licensing public sector information under the Open Government Licence (OGL), and increasing
amounts of public sector information are being made available as open data under this licence,
even by organisations that have traditionally charged for data. For example, Ordnance Survey has
released 11 datasets as open data, there are over 6.000 downloads of Land Registry’s linked data
each month and the Met Office and Companies House have also each opened up some of their
data. There is therefore a visible trajectory towards a world in which increasing volumes of high
quality data are available at low cost or for no charge to users.
Nonetheless, some groups of stakeholders have argued that if the OGL were used for all public
sector information, this could substantially increase the openness of the UK’s public sector
information and could remove many of the barriers to use and re-use that currently inhibit the
realisation of the full value of public sector information.
The Open Data User Group, as a strong advocate of the OGL, has noted that “it needs to become
widely adopted to make the right to data a reality. Conversely, if publishing under the OGL does
not become the default action for public bodies, the right to data will remain an aspiration.” 161
Indeed, it can be argued that should the OGL become the standard licence across all public sector
bodies, this could substantially increase the openness of the UK’s public sector information and
could remove many of the barriers (both in terms of charges levied and restrictive copyright and reuse conditions) that currently inhibit the realisation of the full value of public sector information.
However, others have argued that the public sector information landscape is characterised by its
diversity, and for this reason a ‘one-size-fits-all’ solution is unlikely to be practical or necessarily
desirable. For cases where the release of information under the OGL is not considered appropriate
and cost recovery is justified, the introduction of a generic charging licence could, in principle,
address the current complexity of charging arrangements for public sector information, completing
the UK Government Licensing Framework and simplifying licensing arrangements for users. Some
161
Open Data User Group response to the consultation on the code of practice for datasets and beta charged licence
(http://data.gov.uk/sites/default/files/ODUG%20datasets%20CoP%20response.pdf.pdf)
127
Market Assessment of Public Sector Information
stakeholders pointed out that by having a charge, users and re-users can be re-assured of the
quality of data and be able to expect a certain level of service (this is discussed in more detail
below). However, one concern raised was that such a licence could encourage some PSIHs who
are not covered by the OGL to charge for certain datasets.
Insights:

Making public sector information available under the Open Government Licence is
seen by a large number of stakeholders as an effective means of removing barriers to
exploiting the value of public sector information.

If there are instances where a generic Government Licence for charging for public
sector information are applicable, such a Licence would need to be drafted in a way to
avoid incentivising charging when there is no strong justification or leading it to
become the default alternative to the Open Government Licence.
Economic barriers
The evidence reviewed suggests a number of key economic barriers around maximising the value
from public sector information in the UK. Primarily these relate to the value and cost of datasets,
which can be grouped into two key themes:

which datasets are made available to the general public; and

which datasets consumers have to pay to access for (funding models).
Data gaps
There is currently no national or local information asset register covering the amount and type of
public sector information datasets held 162 (irrespective of whether it is being released to the public)
or the detailed costs of collection and dissemination. Similarly, there is little evidence as to how
consumers themselves are using and re-using particular datasets.
While this has meant this report has been forced to make a number of assumptions in order to
reach quantitative estimates, there is a more significant impact on PSIHs themselves. By not being
able to accurately ascribe value to different datasets, PSIHs are generally unable to reach
evidence-based decisions as to which datasets to publish, how to publish them and what support
to provide. One consequence of this is that the costs (which are readily measurable a priori) of
making particular public sector information datasets available may appear to be much larger than
the benefits – as the benefits case cannot be clearly set out due to a lack of evidence.
Indeed, as part of this report, a sample of government officials have been consulted over the
reasons why public sector information datasets are not released: over 50 per cent of those
responding identified resources as an issue preventing the release of more public sector
information, with the following response being typical: “in a time of diminishing resources and the
need to make best use of the resources we have, the time and the cost of ensuring data validity
before release is an issue.” This was one of the most common reasons given for why data is not
released, but not the only one – others include data protection and legislation.
Stakeholders have identified particular datasets (such as an aggregate Energy Performance
Certificate database) that are currently not available to them (either for free or for a fee) that they
could use to generate new products and services. By not making these datasets more widely
available (or articulating the reasons they are not available), PSIHs may, whether knowingly or not,
be acting as a barrier to realising the full potential of public sector information.
There are various routes to address these data gaps. One route might involve conducting a public
sector-wide audit of public sector information (which also covers the customers using and re-using
162
Though the UK is not alone in this respect.
128
Market Assessment of Public Sector Information
it). This would involve both primary research to determine the stock of public sector information
held across all PSIHs and also identifying how this data is being used commercially across
different economic sectors and actors. This would help match costs and benefits to individual
datasets and building this report to identify those datasets that can generate the most value for the
UK. The costs of doing such an audit would not be insignificant and for it to be effective, it would
need to be repeated in future years (though the on-going costs may be lower than the one-off startup costs). However, the benefits of doing such an audit could be outweighed by the benefits of
PSIHs having a much better sense of which datasets are creating the most value which can guide
future policy decisions around releasing data.
Other routes to addressing data gaps might involve better tracking of usages (perhaps through
measuring how often data is re-used in other sources), requiring all new public sector information
datasets to be logged centrally, or crowdsourcing an audit across the public sector. Another
potential route for the long-term, at least conceptually, might be a true single conduit for the access
of all public sector information.
The Open Data User Group (ODUG) has been set up as an advisory group to Government and
provides a channel through which the potential users of data, from all areas of the community, can
identify their need for data and set out the benefits they expect this data to deliver. This will help
PSIHs understand in more detail the potential benefits which the release of individual datasets can
be expected to deliver.
Through the completion of data gaps, it would be possible for government to articulate what is
meant by the term core reference dataset (beyond the definition contained in the Open Data White
Paper). The issue of core reference datasets is also raised in recent work by APPSI on the national
information framework for public sector information and open data. 163 Consideration could include:

the features of a particular dataset that make it a core reference dataset;

the funding arrangements of core reference datasets (see below);

the obligations of suppliers of core reference datasets; and

the rights of citizens to core reference datasets.
Insights:

There are significant data gaps when it comes to public sector information. This lack
of data can, in some cases, lead to inertia with certain public sector information
datasets not be released or conversely, undue attention being given to datasets that
are unlikely to generate significant value but have a low cost of dissemination.

There are a number of routes to addressing these data gaps. These range from a
detailed, regular audit of public sector information to improved tracking of current
usage.
Funding models
Related to the question of making the business case for more public sector information being
released is the issue of how the cost of generation, collection, retention and dissemination of the
datasets should be funded. This also leads to questions around pricing for access to public sector
information.
This is a complex issue and data on costs (and benefits) is not readily available. Further, one must
take account of different costs being incurred across the different stages of the data lifecycle. For
example, in the case of so-called ‘exhaust’ or ‘by-product’ data that is generated as a result of
163
The issue of core reference datasets is also raised in recent work by APPSI on the national information framework for
public sector information and open data, www.nationalarchives.gov.uk/documents/nif-and-open-data.pdf.
129
Market Assessment of Public Sector Information
PSIHs conducting their day-to-day and other activities and duties, the marginal cost of its
generation will, by definition, be negligible or very low compared to the activity that caused the data
to come about; but the marginal cost of its dissemination to the wider public may be higher and
vice versa for some ‘purposely collected data’. There will also be additional costs in formatting the
data for use and further costs if support is provide to the public in re-using the data.
The issue of charging and funding models for public sector information is highly contentious and
extremely complex. For example, there remains an element of subjectivity as to what constitutes a
dataset and what constitutes a ‘value-add’ service – with some PSIHs arguing that what is being
charged for is not the public sector information but its interpretation and analysis.
As part of this report, a wide range of opinions have been expressed as to whether there is any
economic or other justification for charging for public sector information datasets. On the one hand,
these arguments include:

charges for datasets create barriers to entry and expansion for SMEs and individuals to
develop new products and services;

the charges prevent SMEs and individuals from ‘experimenting’ with the datasets before they
purchase to see if they are able to derive value from them, thereby making it hard to develop
business cases; and

any lost revenues to PSIHs from releasing datasets for no cost will be recovered by the
Exchequer in the long-run through increased tax revenues and more jobs being created.
In contrast, arguments have been put forward to support current pricing arrangements include:

aligning a revenue stream with a particular dataset will ‘protect’ it from any reductions in
funding, allowing PSIHs to continue to supply this even if they themselves must make other
savings;

a price can be interpreted as a signal of consumers’ willingness to pay for a particular dataset’s
quality and a commitment by the PSIH to maintain this and offer support; and

charging for certain datasets is necessary given they include elements of commercial or
international datasets.
As the most visible PSIHs that charge for certain public sector information, a great deal of attention
has been focused on the four Trading Funds that make up the Public Data Group (PDG):
Ordnance Survey, the Met Office, Land Registry and Companies House. The view has been
expressed by some start-up companies that these PSIHs hold some of the most valuable datasets
and there are strong arguments that these should be treated as core reference datasets available
to all at no direct cost to the general public.
It is not helpful to treat these PSIHs as a single group – indeed, there are a number of other public
bodies that charge for access to datasets 164 and there are a number of other Trading Funds
outside the PDG. The four Trading Funds in the PDG differ in their sources of revenue, their public
service duties and whether the public sector information they hold can be classed as ‘exhaust’ data
or ‘purpose-collected’ data. Further, it should be noted that substantial progress has been made by
the four Trading Funds in recent years to make increasing volumes of data available as open
data 165 and there continue to be moves in this direction.
However, despite these positive steps, there remains a perception among many consumers and
commentators 166 that they are unable to access certain datasets for reasons of cost and this is
creating a barrier to business growth. A number of studies (e.g. Pollock, 2011 and others – see
164
For example, the Office of National Statistics has a subscription charge for certain datasets.
For example, see www.ordnancesurvey.co.uk/oswebsite/products/os-opendata.html
166
See for example www.freeourdata.org.uk/ and other references in Appendix 3.
165
130
Market Assessment of Public Sector Information
Appendix 3) have argued that releasing these datasets as open data will have significant welfare
benefits.
The impact of cost recovery model for public sector information, compared to the Open
Government Licence model, is summarised below.
Figure 6.3: potential impact of the cost recovery model on the public sector
information market
Taxpayer funded public sector information
Public sector information supplied on a cost
recovery basis

Public sector information typically made
available at no cost to users under the Open
Government Licence

Access to public sector information involves a cost to
the user and may come with restrictions

There may be strong public policy reasons for
having these datasets available at no cost

The costs of collection (rather than dissemination)
are significant and cannot be borne by the public
purse

Will not include other commercial or international
datasets

May include other data sourced under licence

The availability of public sector information at no
cost to the user can contribute to the
transparency agenda by increasing access to
the widest possible customer base, irrespective
of the ability to pay

Access to Public sector information is restricted on
the basis of cost

Will not typically include other services or any
other guarantees

May also include bespoke value-add services and
guarantees of data quality and continuity
Source: Deloitte analysis
It is very difficult to perform a robust cost-benefit analysis of different funding model options for
PSIHs, not least because PSIHs differ greatly from one to the next. While some data is available
as to the costs incurred from collection and dissemination, this is not typically openly available or
apportioned by dataset, nor is data readily available on how their customers are using the datasets
and generating revenue and value.
More fundamentally, very little data is available on what the benefits might be if charging models
were to radically change – as it is very difficult to predict how businesses and individuals might use
datasets in the future to generate new products and services, and by implication impact economic
growth. It is also important to note that, as per the HMT Green Book 167 guidance, any benefits from
a change in charging structures should include not just increased tax receipts but wider social
benefits and costs in terms of organisational impacts.
Even in the case of the four Trading Funds that make up the PDG, estimating the effects on
Exchequer revenue of releasing all their public sector information as open data is a difficult task, in
spite of the information made available by the Trading Funds as part of the study. There are
differing views on precisely which revenues should be considered relevant to PSI, considering
factors such as specific revenues from the data, the cost of collecting the data, the extent to which
Government ‘buys’ data from itself and so forth.
That notwithstanding, on the basis of the information made available to this study, it is possible to
indicatively calculate that the cost effects on Exchequer revenue of continuing to collect and
disseminate Trading Funds’ public sector information in its current guise without charging for it.
This cost is estimated to be in the order of £395 million on an annual basis 168 . However this figure
167
168
Available at www.hm-treasury.gov.uk/data_greenbook_index.htm
This figure comprises of Trading Funds’ operating surpluses and the cost of data collection.
131
Market Assessment of Public Sector Information
is without regard to the extent that Government pays for public sector information from the four
Trading Funds in the PDG (this varies significantly between Trading Funds). On the basis of
information provided on sales channels, it can be estimated that the annual loss to the Exchequer
would be lower, as government would no longer need to purchase these public sector information
datasets – it could use them at no cost. In this scenario, the loss to the Exchequer on an annual
basis might be of the order of £143 million. This figure may be lower still if there are efficiency
savings to be made if fewer dedicated sales and marketing resources are required by Trading
Funds.
Following the HMT Green Book approach to account for the wider social and economic benefits, it
is important to note that in a world without charging, private sector entities (consumers and
businesses) that currently pay for access to public sector information provided by the Trading
Funds would benefit by this amount - around £143 million – and some of this may be recouped in
the medium-term as a result of additional economic activity generating tax revenues.
An additional group that is currently deterred by having to pay for the data would also benefit as
they are able to access the data. Estimating the size of this latter group is difficult but directionally it
is clear that removing charging would mean more people and businesses, not fewer, would be able
to access and benefit from this data 169 . Conversely, organisations who are at an advantage in
using their own proprietary information for commercial advantage, might find their competitive
advantage eroded if more public sector information is released to act as a publically available
substitute to that data.
The situation is thus complex, and while the above example is stylised, it does suggest the
quantum necessary for the associated benefits to outweigh the costs. Without more detailed and
accurate data on both costs and benefits (including wider social benefits) it is not possible to reach
a clear conclusion on this issue.
However, this is not to say there is no room for improvement today in the provision of public sector
information that carries a charge and a number of steps are being taken in this respect. These
include:

much more communication about what existing licences allow consumers to access and use /
re-use the public sector information for, building on existing efforts by the trading funds and
other PSIHs to build awareness among the user community;

offering substantial discounts to SMEs and individuals;

implementation of a ‘royalties’ model for consumers to exploit the value of public sector
information up to a certain value before a charge is applied;

greater provision of ‘sandbox’ or secure environments in specialised locations across the UK to
allow consumers to explore datasets;

greater provision of out-of-date public sector information at no cost to the general public to
allow consumers to experiment; and

greater use of ‘hack days’ to demonstrate the value of particular public sector information
datasets.
These options have not been costed or their benefits estimated.
A number of these initiatives are already being pursued and there appears scope and benefit to
their wider roll-out. Of course, not all of these initiatives will be applicable to all PSIHs that charge
for public sector information – there is no ‘one-size-fits-all’ model – but they should work with the
grain of the market and build on existing initiatives to release more data. In some cases there may
be significant logistical or legislative challenges to overcome before the above suggestions are
implemented. For example, under the IFTS differential charging is prohibited; and a single sandbox
169
There would accordingly be second-round effects for people consuming products provided by these businesses, etc.
132
Market Assessment of Public Sector Information
model may not effectively meet the needs of all developers. There are also questions whether
PSIHs can effectively separate communication and marketing efforts promoting the use of their
public sector information from marketing to promote their value-added services which may be in
direct competition with the private sector.
Insights:

The issue of charging for public sector information datasets and their funding models
is complex. While significant progress has been made by a number of major PSIHs in
this area in simplifying charging structures and making more public sector
information available as open data, there remains a perception that barriers exist and
there is scope for improvement.

Ways to improve access could include greater communication on licence conditions,
discounts to certain consumers, a royalties model, use of a ‘sandbox’ model, greater
provision of out-of-date information and more hack days.

There is currently a lack of data to definitively conduct a cost-benefit analysis across
different funding models across the range of PSIHs currently charging. This is an area
for further analysis through primary research with the direct customers of public
sector information and a detailed cost apportionment exercise to assign costs to
individual datasets.
Access barriers
The evidence reviewed suggests a number of access barriers to fully maximising the value of
public sector information:

difficulties around finding where public sector information is located;

a lack of skills and understanding to fully exploit public sector information;

the format and reliability of public sector information; and

a reluctance to use and share public sector information.
Public sector information fragmentation
Fragmentation in the supply of public sector information continues to be a problem for many
consumers, even following the establishment of a number of data portals. A sample review of
websites done as part of this study has found that too often it is difficult to locate a clear point of
contact, or establish who has ownership of and responsibility for a particular dataset. Even on
data.gov.uk, where contact details are provided, these are often generic enquires email addresses
rather than named contact email addresses.
In contrast, this report has received feedback that when consumers have found the relevant point
of contact in a PSIH their experience has been very positive with a productive dialogue being
established. Indeed, some individual PSIHs have established clear procedures for consumers
seeking to raise queries or challenge restrictions, although this is not yet happening in all areas of
the public sector. Dialogue can also take place between users: data portals such as data.gov.uk
and the London Datastore have busy forums and request pages which encourage an active
dialogue between information users and re-users and information holders.
Public sector information fragmentation can, at best, raise transaction costs from dealing with
public sector information and, at worst, deter users and re-users from using public sector
information altogether, or make this impossible to achieve. Reducing fragmentation across the
133
Market Assessment of Public Sector Information
public sector could save up over £50 million per annum in terms of reduced transaction costs and
time saved for data specialists 170 .
However, it should be noted that the existing fragmentation and opacity of some public sector
information datasets has created market opportunities for some intermediaries to develop new
products that aggregate datasets and present it innovative ways. Thus, PSIHs need to consider
who is using their data and how and whether there is a risk of crowding out innovation if they
intervene to reduce fragmentation in such a way that duplicates what the private sector can
provide.
Insights:

Improvements are on-going to reduce the time taken to locate public sector
information datasets. Reducing this can help reduce transaction costs and the overall
cost of doing business.
Data scientists and the skills gap
A number of studies have recently been published contending that advanced economies face a
skills gaps in so-called ‘data scientists’. Although statisticians and experts on quantitative analysis
have long existed, data scientists differ from these existing professions in a number of important
ways. As well as being able to work with large volumes of structured and unstructured data, they
are able to translate these analyses into policy and commercial-ready insights and effectively
communicate them to a range of stakeholders, often using innovative tools and visualisations. Key
to this is the ability to “identify rich data sources, join them with other, potentially incomplete data
sources, and clean the resulting set.” In many ways they therefore resemble scientists more
closely than traditional data analysts.
There is a fear that a lack of data scientists will reduce the UK’s competitive advantage. The
evidence received suggests that in the UK:

there is increasing demand for individuals with a portfolio of skills able to manipulate
quantitative data, present it in innovative ways and generate commercial and policy insights
from it;

many of the individuals performing these roles have no specialised training, but rather have
learned on the job and / or have a science/computation/mathematics background;

businesses rarely designate specific ‘data scientist’ roles; rather, such analyses are done
across a combination of professions such as statisticians, economists, researchers, analysts,
policy and commercial managers – a dedicated data scientist would embody elements of all
these roles;

certain industries such as pharmaceuticals, financial services, professional services and retail
are increasingly dependent on these skills sets and a shortage of them would reduce the UK’s
international competitive advantage.
Data scientists will also be conversant with the vocabulary of public sector information and open
data and have the skills to create and manipulate large datasets and linked data.
Based on ONS figures for 2011, around 1.5 million workers in the UK, representing around 5 per
cent of the active workforce, are employed in job categories that are likely to involve elements of
the role of the data scientist, but which individually may not be termed data scientists. The average
annual median wage of these workers was over £36,000 in 2011 which is higher to the national
annual median wage of £26,000.
170
Based on an assigned average value of time spent by individuals in occupations that regularly use public sector
information.
134
Market Assessment of Public Sector Information
Economic theory suggests that a skills shortage will manifest itself in the form of large wage
differentials between ‘data scientists’ and other comparable professionals. A recent report found an
observable pay premium for ‘big data’ staff in 2012, with salaries around 20 per cent higher than
those for IT staff as a whole. 171 This may persist due to lags between training and entering or reentering the workplace, but economic theory suggests it will eventually dissipate in the long-run as
supply increases to meet demand and is able to exploit the full value of public sector information.
Evidence suggests the market is beginning to address skills shortages 172 through:

sector specific leading practices, lowering complexity and the subsequent talent learning curve
(including setting legal precedents for legal liability and compliance); and

universities are having new certifications for big data disciplines.
However these initiatives will take time to work through the system with the interim consequence
being that some value of public sector information may remain locked up. The general scarcity and
increasing competition for these skilled workers from the private sector makes it harder to construct
the infrastructure for world class public sector information. A shortage of data scientists also
hinders efforts to scale-up public sector information data analytics.
Within the public sector, concerns have been highlighted over a lack of skills and familiarity to work
effectively with data. These concerns should not be overstated as public sector officials have a
long history of using public sector information to inform policymaking without having dedicated data
scientists. What the concerns appear to be directed at are cultural biases against using public
sector information from outside home PSIHs, as well as having the necessary skills to combine
and manipulate Big Data and Linked Data 173 . Comments from stakeholder conversations and the
government official’s informal consultation reinforce this concern – the following being
representative:


“There is currently a lack of awareness and skills in blending and combining multiple
datasets. This will need to be addressed if we are to fully exploit the power and potential of
integrating disparate sources over the web.”
“Data users do not fully appreciate the power and potential of open data. There may also
be cultural resistance to change:, for instance, through not trusting or being able to exploit
new, untried and untested third-party datasets in analysis and policy advice.” 174
Where public sector employees do not have a numerate or scientific higher education background,
or are not accustomed to working with large datasets, they could benefit from increased training
and support in this area. However, there are understandable concerns over resources at a time
when departmental budgets are under considerable pressure. The danger, therefore, is the training
in this area is viewed as a ‘nice to have’ which is not currently a budgetary priority and is therefore
neglected. Were this to happen, much of the effort to open up public sector information and
facilitate sharing across the public sector would be poorly spent, as public sector employees would
lack the skills to exploit the available data – the PSIHs would only be able to operate on the supply
of the market, lacking the skills to be innovative public sector information consumers.
171
See e-skills UK, ‘Big Data Analytics: an assessment of demand for labour and skills, 2012-2017’ (January 2013)
See Deloitte Tech Trends 2013, available at
www.deloitte.com/view/en_GB/uk/services/consulting/technology/technologytrends/abbffbfdad4ac310VgnVCM3000003456f70aRCRD.htm
173
However, it should be noted that many cases the linking itself is carried out by specialist data companies, which may
be an example of increased activity in this space creating or expanding a market for private companies. This minimises
the internal skill base that public sector bodies are required to develop, reducing the burden of releasing data: they are
able to release data in raw form for linking by third parties. Nonetheless, some public sector bodies which work
extensively with data, such as HEFCE, are known to have developed in-house capabilities in this area. In addition some
researchers in the field, recognising that they depend heavily on linked datasets, have developed the technical expertise
to do this work themselves.
174
Government officials consultation, November 2012
172
135
Market Assessment of Public Sector Information
There are a number of routes around these budgetary concerns. One example is the use of
massively open online courses (MOOCs) to enhance skills sets and increase expose to data
manipulation techniques. These courses are free of charge to the user and include offerings from a
number of prestigious higher education institutions. For a limited time commitment over a period of
weeks, generally between three and six hours, students receive online material from academics,
as well as access to a range of study material, online forums, and in some cases certification
following course completion and assessment. Using estimates of public service wage levels and
assuming an eight week long course of six hours of study a week, the opportunity cost value of
taking a MOOC during office hours is in the order of £500 per participant, although the benefits
from improved skills are likely to be much higher.
In practical terms, measures could be taken via internal communications to raise awareness of the
benefits of taking these courses. In addition it would be helpful if completion and accreditation of
relevant courses was viewed favourably in performance assessments, which would act as an
incentive. A commonly used approach to in transformation programmes is to identify ‘champions’
across all grades who can be encouraged to take the courses and then act as advocates among
their colleagues; once a critical mass of understanding and enthusiasm is attained, this becomes
self-reinforcing and employees should start taking the courses as part of their continuing training
and development.
A selected list of relevant MOOCs, available at the time of writing, is included below for
guidance. 175
Course name
MOOC
provider
Higher
Education
Institution
Estimated
time
commitment
Web link
Passion driven
statistics
Coursera
Wesleyan
University
3-4 hours per
week (6
weeks)
www.coursera.org/#course/pdst
atistics
Web intelligence and
big data
Coursera
IIITC
2-3 hours per
week (10
weeks)
www.coursera.org/#course/bigd
ata
Introduction to data
science
Coursera
University of
Washington
10 weeks
www.coursera.org/#course/data
sci
Statistics: making
sense of data
Coursera
University of
Toronto
6-8 hours per
week (8
weeks)
www.coursera.org/#course/intro
stats
Data analysis
Coursera
John Hopkins
University
3-5 hours per
week (8
weeks)
www.coursera.org/#course/data
analysis
175
Note: the above is a selective list, and is not intended to be a comprehensive survey of all relevant MOOCs.
Availability of the above courses may alter over time as the majority of the courses listed are to commence in Spring
2013. The level of suggested background knowledge varies, and not all of the above courses will be appropriate for all
employees. We have selected the courses based on a combination of our personal experience and our assessment of
the range of data analysis needs that may be useful for a public sector employee, but we are unable to offer any
comment or assurance as to the quality of the course content or delivery, or the continuing availability of any of these
courses.
136
Market Assessment of Public Sector Information
Course name
MOOC
provider
Computational
methods for data
analysis
Coursera
Computing for data
analysis
Coursera
Estimated
time
commitment
Web link
University of
Washington
10 weeks
www.coursera.org/#course/com
pmethods
John Hopkins
University
3-5 hours per
week (8
weeks)
www.coursera.org/#course/com
pdata
Higher
Education
Institution
Note: no analysis has been done as to the quality of the above courses’ content, which could vary significantly.
Traditional training will of course continue to play an important role, as well as interactive and
workshop sessions, such as mash-up days, especially those involving external developers. These
are useful for sharing knowledge and expertise and creating an environment which is conducive to
experimentation and innovative thinking.
In addition, it may be that public sector organisations will wish to increase their recruitment of those
with a scientific or numerate higher education background, so as to expand the pool of relevant
talent upon which they are able to draw. This will vary according to the organisation in question
and the appropriate skills mix for the performance of their public task; for this reason we do not
make an across the board recommendation regarding recruitment. However, it is likely to be
helpful for public sector organisations to consider the alignment of their current and future skills
based with the increasing availability and use of information from across the public sector.
There will also be a role for Cabinet Office and other public sector organisations to build on notable
achievements 176 in public sector information provision to foster a more conducive culture within the
public sector to using public sector information innovatively. The benefits accruing from improved
use of information within the public sector are likely to outweigh these. Adapting the McKinsey
analysis used by Policy Exchange 177 , improvements in efficiency of between 1 and 5 per cent can
lead to annual savings of between £1 and £8 billion nationally and around £70 million in local
government (on a smaller savings ratio).
Insights:

While there may be gaps in the supply of ‘data scientists’, economic theory suggests
that in the medium- to long-term, the number of data scientists will increase, filling the
supply gap and reducing the current wage premium.

However, in the short-term, this may mean public sector information is left underexploited and value remains locked. The general scarcity and increasing competition
for these skilled workers can make it harder to construct the infrastructure for world
class public sector information and scale up efforts to exploit its value.

There are some low-cost solutions that can be explored to quickly improve the skills
base to be able to effectively manipulate and extract value from public sector
information.
176
These include (i) a process to drive the supply of open data from Whitehall, (ii) achieve the release of over 9,000
datasets on data.gov.uk, (iii) assist in the establishment of the Open Data Institute; and (iv) support the UK’s efforts
globally to improve open data and the wider transparency agenda of the Open Government Partnership and the G8.
177
See www.policyexchange.org.uk/publications/category/item/the-big-data-opportunity-making-government-fastersmarter-and-more-personal
137
Market Assessment of Public Sector Information
Format and reliability
The release of data in an unfinished form raises concerns for the public sector too because of
concerns that the public may be provided with data that is inaccurate and potentially misleading,
which can have negative consequences.
With respect to the format of public sector information, its importance may vary between
customers. For casual consumers, it is clearly of the utmost importance, and they may most value
format. In contrast, professional users may be more concerned with consistency of service,
commitment to on-going supply of data and data and can accommodate changes to format.
Equally, intermediaries may positively value low quality data as they can provide a service to
improve the quality and format of public sector information for wider consumption.
Evidence as to how significant a barrier this currently forms to the use and re-use of public sector
information is mixed. In general, there seems to be steady improvement in all the areas of format,
although there remains work to be done – especially with regard to ensuring consistency in format
and upgrading the star rating of datasets. Progress is also being made with data being updated
more consistently. Examples of this include data released by the Met Office, and transport data
released directly by Transport for London as well as through data.gov.uk.
Land Registry’s release of PPI data is an example of good practice. The data was released in CSV
(3 star format) then upgraded to linked data (4 star standard). The quality of the data was also
added the organisation’s list of Key Performance Indicators and anyone publishing the data is
asked to include a link to a dedicated Land Registry team for reporting any inaccuracies in the
data.
With respect to format, it is well established that data which is released should, wherever possible,
be in machine-readable formats. Format is therefore linked to usability. For a commercial service to
be based on an data source, there must be confidence that the data will be maintained and kept up
to date. The ODI notes that “the sustainability of data publication is all important for those who
build services on top of that data. So long as it is machine-readable (2-star) and openly
licensed…the frequency, consistency and accuracy of the publication of a dataset is more
important than the format in which it is published.” 178 The reasons for this are easy to understand:
if a business offers a service to customers, it cannot afford for that service to be unpredictably cut
off or to offer out of date information. This is especially true of time-sensitive (volatile) information,
such as travel updates and weather forecasts.
As highlighted in an earlier chapter, the working assumption of this report is that around 50 per
cent of datasets are three stars or above (based on a review of data.gov.uk datasets and other
sources including the data released by local government). The question is then whether this
current level of usability acts as barrier to use and re-use of public sector information.
Conceptually, datasets that fail to meet the required standard may be constraining the ability to
generate value from public sector information datasets in a number of ways:

requiring users and re-users to have specialist propriety software to view public sector
information datasets. Examples include the early release of the COINS database, which due to
its size and nature was difficult for users to manipulate;

preventing users and re-users from exploiting the benefits of semantically linked datasets; and

restricting access to certain public sector datasets and limiting their uses / re-uses.
Five star data offers significant advantages to users in terms of combining datasets together, which
is one of the areas where this report has identified large potential benefits to be generated.
However, moving towards a five star standard could involve significant cost to PSIHs. Following
discussions on the merits of having all data meeting this standard with members of the start-up
178
See www.theodi.org/consultation-response/improving-local-government-transparency
138
Market Assessment of Public Sector Information
community and other stakeholders, it appears that while it is considered desirable, the consensus
seems to be that as long as data is machine-readable the lack of a five star rating is not currently a
significant barrier for users and re-users. There is concern that pushing too hard for all data to be
upgraded to five star status could dissuade information holders from releasing data at all, given the
cost. In general, users have expressed a preference for data to be released earlier with a lower
star rating, and then upgraded when possible.
In addition, consistency of format across comparable datasets is important, as this affects how
easily they can be compared and combined. This is a particular issue where there are many
organisations with their own outputs, such as local authorities and the NHS.
Of course, as noted earlier, a higher star rating cannot be viewed as a simple proxy for higher data
quality. A dataset may be in a machine readable format and available as linked data, but still
contain inaccurate or incomplete data. In general, stakeholders have reported that the quality
(including reliability of delivery [see below] and accuracy) of the data is of a higher priority that
upgrading it beyond two or three stars, since the quality affects the fundamental usefulness of the
data.
The Open Standards Principles 179 , the Standards Hub 180 , the Open Standards Board and the
public sector will be crowdsourcing, researching and implementing data standards for Government
IT systems. The user challenges it will focus on are likely to cover standards relating to formats
and meta-data that should help to provide data on reliability. Some of these open data standards
may be made compulsory for central government use.
However, it should be noted that a minority of stakeholders have complained that datasets
released by different PSIHs are not always easy to combine and work across, because of
variations either in the content or the format. This is a particular issue with data produced by local
authorities, with each local authority often adopting their own standards and procedures. This
report recognises that there are on-going efforts, such as e-PIMS, to tackle this issue and secure a
greater degree of standardisation across PSIHs.
Further, there appears to be scope for improvement is greater certainty and clarity over the
publication schedule of public sector information datasets and what users can expect from PSIHS.
In particular, certainty for businesses making investment decisions to use public sector (and other)
information datasets could be improved through:

having a clear articulation of each dataset’s publication schedule;

having a cover sheet setting out the limitations of the data, explaining outliers and providing
links to previous analyses (greater use of metadata 181 );

having a clear indication of when any given dataset may be discontinued; and

the level of PSIH support that will be provided (perhaps at a charge).
The levels of support and explanations of the dataset could vary by dataset type.
Clearly there will be a cost involved in this for PSIHs (one-off and on-going), but the benefits to
businesses from increased certainty could be significant.
Insights:

Improvements continue to be made on the quality, format and consistency of public
sector information.

There is scope for improvement, especially around the greater provision of metadata.
179
See www.gov.uk/government/news/government-bodies-must-comply-with-open-standards-principles
See http://standards.data.gov.uk/
181
That is, broadly speaking, data about data.
180
139
Market Assessment of Public Sector Information
Chapter summary

This chapter has summarised the main barriers to the release, use and re-use of public sector information.
These include
o
Legislative barriers: while current legislation, guidance and regulations are not hindering the public
sector information market themselves, there is often a perception (wrongly or rightly) that some
legislation, guidance and regulations prevent datasets from being released and shared thereby
reducing the availability of datasets. The Open Government Licence is widely seen as an effective
means of improving the availability of public sector information;
o
Economic barriers: there are significant data gaps around exactly what public sector information
exists, making it difficult to reach decisions around its optimal provision. The issue of charging and
funding models for public sector information is complex, with the lack of data making it hard, at
present, to accurately conduct cost-benefit exercises on the extent that charged-for data might be
delivered for free in future; and
o
Access barriers: improvements continue to be made to reduce fragmentation and improve the
consistency of public sector information; however there remains scope for improvements in the
accessibility of public sector information datasets to the general public. Equally, in some parts of
the private and public sector, there is a skills shortage preventing the effective extraction of value
from public sector information.
140
Market Assessment of Public Sector Information
7. Conclusion
This report has considered, for the first time, in its entirety, the public
sector information market in the UK. It has provided details on the
supply and demand sides of the market, the legislative and regulatory
framework, the current value of the market and identified a number of
barriers that are preventing the UK from maximising the full value of
public sector information. This report forms the evidence base for the
Shakespeare Review, which is making recommendations to
Government.
This concluding chapter sets out some closing thoughts on the subject
of public sector information in the UK.
Closing thoughts
141
Market Assessment of Public Sector Information

Through a review of the literature and real-life examples of its use and re-use, it is clear that
public sector information is generating significant value for the UK currently – in terms of
financial, economic value and employment benefits, but also wider social benefits.

The increasing availability and better quality of public sector information can lead to even
more value being generated in the UK and also overseas through network effects.

By being a global leader in the release of public sector information, the UK government is
helping create an environment that is conducive for innovation and the development of new
skills that could give UK businesses a competitive edge overseas 182 .

This report has highlighted that while the availability and quality of public sector information
has improved in recent years, which has helped the market evolve and grow, there remain
certain barriers that may hinder or constrain the continued growth of the market. The policy
insights raised suggest that, in some cases, there may be a role for Government to step in
and address market failures.

In addressing market failures, Government will need to balance the interests of different
stakeholders, but equally it should not necessarily be constrained by the current fiscal
environment in order to consider a much longer time-horizon of benefits and costs.

However, it should not be forgotten that public sector information is only one component in
the wider data landscape – the increased availability of more and better private sector data
(open or otherwise) will also have significant benefits.

Indeed, a key barrier identified has been the lack of information on how businesses use
public sector information. Fostering a culture of greater openness and collaboration around
data and its use will benefit all parties. For example, one way of collaborating and fostering
more openness in future, providing it can be carried out in a not-too-onerous manner, might
be providing data free to third parties on the pre-condition that the information flow becomes
a two-way street for Government and policymakers, e.g. “you can have our data, but we’d
like to know how it is being used and re-used to give us insight, benefit us and in turn UK
society”.
182
Though this has not explored in detail as part of this report.
142
Appendix 1: Glossary
These glossary definitions have been taken from a variety of sources including the Open Data
White Paper 183 (June 2012), the Advisory Panel on Public Sector Information (APPSI) Glossary
website 184 and the OFT Commercial use of Public Information report 185 (the ‘CUPI’ report,
December 2006). As noted in that document, given the relatively nascent nature of the market and
rapidly changing landscape, these definitions will have an element of. At the time of writing
(January 2013), Cabinet Office is consulting to gain a collective view on definitions to be used in
forthcoming Transparency and Open Data publications. This is to be done via a ‘wiki’ site hosted
by the APPSI 186 .
For reference the source for each glossary term is listed. Where there is no term this refers to
cases where the term has been defined specifically for this research.
For reference the source for each glossary term is listed. Where there is no term this refers to
cases where the term has been defined specifically for this research.
Aggregated data
A form of anonymisation of unit records involving combinations
such that individual records are not disclosed.
Source: APPSI Glossary
Anonymised data
Data that has been adapted so that individual businesses,
individuals and other organisations cannot be identified from it.
Source: adapted from APPSI Glossary
Application programming
interface
A specification intended to be used as an interface by software
components to communicate with each other. An API may include
specifications for routines, data structures, object classes and
variables.
Source: adapted from APPSI Glossary
Attribution licence
A licence that requires that the original source of the licensed
material is cited.
Source: Open Data Handbook
Asset list
A register of data and information items held by a PSIH which are
of interest or value to the PSIH itself, and potentially to others.
Source: OFT CUPI
Authoritative
Able to be trusted as saying something. Some information can be
accurate but not authoritative, but it will be both if it comes from a
source with authority to provide it.
Source: APPSI Glossary
Big Data
Gartner 187 describes Big Data as data defined as high volume,
velocity and variety information assets that demand cost-effective,
innovative forms of information processing for enhanced insight
and decision making.
183
Available at: www.gov.uk/government/publications/open-data-white-paper-unleashing-the-potential
See www.nationalarchives.gov.uk/appsi/appsi-glossary-a-z.htm
185
Available at: www.oft.gov.uk/OFTwork/publications/publication-categories/reports/consumer-protection/oft861 186
See www.nationalarchives.gov.uk/appsi/open-data-psi-glossary-pilot.htm. At the time of writing the wiki had not yet
been established.
187
See http://www.gartner.com/it-glossary/big-data/
184
Market Assessment of Public Sector Information
Source: Gartner
By-product data
Data that is generated through the performance of regular or oneoff activities, where the generation of the data was not the primary
objective of the activity or part of its Public Task. Also referred to
as ‘exhaust data’.
Class licence
A licence that sets out standard terms and obligations, enabling
the re-use of a particular class or category of material.
Source: OFT CUPI
Click-use
The online licensing system for Crown and Parliamentary
copyright information developed by the Office of Public Sector
information in 2001. This has subsequently been superseded by
the Open Government Licence and Open Parliament Licence, but
remains historically significant.
Source: UK Government Licensing Framework
Commercial use / re-use
Use or re-use that is intended for or directed towards commercial
advantage or private monetary compensation. For clarity,
commercial advantage applies either to reselling the data through
products in any form or to internal use or re-use to improve
business effectiveness.
Source: The National Archives Information Management
Glossary with APPSI Glossary addition
Consumer surplus
The value or benefit consumers of a product or service enjoy over
and above the price they paid for it. It is the difference between
the price consumers pay and the price they are willing to pay.
Copyright
Part of the family of intellectual property rights including
trademarks, designs and patents. Copyright applies automatically
when a work is created in a material form. Copyright applies to
literary works, such as website articles/annual reports; artistic
works maps, drawings, paintings and photographs; films; sound
recordings; broadcasts; dramatic and musical works and
typographical arrangements. The first owner of copyright will
normally be the artist/author or organisation that created the work
(except for Crown copyright). Copyright subsists in a work
regardless of the level of artistic or literary merit. The standard
term of copyright is the life of the author plus 70 years.
Source: The National Archives Information Management
Glossary
Core reference data
Authoritative or definitive data necessary to use other information
produced by the public sector as a service in itself due to its high
importance and value.
Source: Open Data White Paper
Crown body
An organisation which acts on behalf of the Crown, meaning the
sovereign acting in a public or official capacity. This includes most
central government departments including government Trading
Funds. In many cases the Crown status or otherwise is specified
within the context of legislation.
Source: OFT CUPI
Crown copyright
Crown copyright covers material created by civil servants,
ministers and government departments and agencies. It is legally
144
Market Assessment of Public Sector Information
defined under section 163 of the Copyright, Designs and Patents
Act 1988 as works made by officers or servants of the Crown in
the course of their duties. Copyright made by Her Majesty or by
officers can also come into Crown ownership by means of an
assignment or transfer of the copyright from the legal owner of the
copyright to the Crown.
Source: APPSI Glossary
Crown copyright waiver
Categories of material on which the Crown asserts its copyright
but waives it and which is not subject to formal licensing or
payment.
Source: The Future Management of Crown Copyright White
Paper
Customer insight Data
Data or information recording users’ accounts of their experience,
with an assessment of public service providers.
Source: Open Data White Paper
Data (singular or plural)
Qualitative or quantitative statements or numbers that are
assumed to be factual and not the product of analysis or
interpretation. Data can be structured or unstructured.
The term structured data refers to data that is identifiable because
it is organised into a recognisable structure. The most common
form of structured data (structured data records (SDR)) refers to a
database where specific information is stored based on a
methodology of columns and rows.
In contrast, unstructured data has no identifiable structure.
The terms data, information and knowledge are frequently used
for overlapping concepts. The main difference is in the level of
abstraction being considered. Data is a broad term, embracing
others, but is often the lowest level of abstraction, information is
the next level and, finally, knowledge is the highest level.
Source: adapted from APPSI Glossary and Open Data White
Paper
Data controller
A person who (either alone or jointly with other persons)
determines the purposes for which and the manner in which any
personal data are, or are to be, processed.
Data discovery benefits
Benefits (economic or otherwise) that arise from the analysis of
data and information itself.
Data enabler
An intermediary that facilitates public sector information or open
data initiatives without actually publishing or consuming data
themselves, usually through software, platform, data centre
infrastructure etc. Service types include platform-as-a-service,
infrastructure-as-a-service, and Software-as-a-service.
Data exploitation benefits
Benefits (economic or otherwise) that arise from using data and
information as inputs into products and services.
Data holder
The public sector body holding public sector information. Also
referred to as Public Sector Information Holders.
Data infomediary
An intermediary that aggregates, scrapes or collects publicly
available data or data stores that enhance the raw data with
visualization, completeness, accuracy, analysis, or accessibility.
145
Market Assessment of Public Sector Information
Data mash-up
Combining different and distinct datasets to create new datasets
and generate new insights.
Data owner
The data owner (organisation or individual) is responsible for
understanding what information is brought into a system,
assigning meanings to data collections and constructing and
modifying data models.
Data processor
With respect to personal data, this refers to any person (other
than an employee of the data controller) who processes the data
on behalf of the data controller.
Data protection
The Data Protection Act 1998 defines the ways in which data and
information on living persons may be legally used and handled.
The Act sets out the fundamental principles with which personal
data much satisfy, including:
(a) be processed fairly and lawfully;
(b) be obtained only for lawful purposes and not
processed in any manner incompatible with these
purposes;
(c) be adequate, relevant and not excessive;
(d) be accurate and current;
(e) not be retained for longer than necessary;
(f) be processed in accordance with the rights and
freedoms of data subjects;
(g) be protected against unauthorised or unlawful
processing and accidental loss, destruction or
damage; and
(h) not be transferred to a country or territory outside the
European Economic Area unless that country or
territory protects the rights and freedoms of the data
subjects.
Source: adapted from Data Protection Act
Data scientist
An analytical role that is built on three core skills: data
management, data analysis and business and policy insight.
While the formal training of data scientists is typically in statistics,
mathematics or computer science, they will also have strong
communication skills and business/policy acumen. They have
been described as “part analyst, part artist” by IBM.
Data sharing
The transfer of data between different organisations to achieve an
improvement in efficiency and effectiveness. This document
assumes that data sharing will continue to operate in line with
current domestic legislation and the UK’s international obligations.
Source: Open Data White Paper
Data subject
Under the Data Protection Act 1998, this refers to an individual
who is the subject of personal data.
Dataset
As defined in the Protection of Freedoms Act 2012: “ ‘dataset’
means information comprising a collection of information held in
electronic form where all or most of the information in the
collection:
(a) has been obtained or recorded for the purpose of
146
Market Assessment of Public Sector Information
providing a public authority with information in
connection with the provision of a service by the
authority or the carrying out of any other function of the
authority,
(b) is factual information which —
(i) is not the product of analysis or interpretation
other than calculation, and
(ii) is not an official statistic (within the meaning
given by section 6(1) of the Statistics and
Registration Service Act 2007), and
(c) remains presented in a way that (except for the
purpose of forming part of the collection) has not been
organised, adapted or otherwise materially altered
since it was obtained or recorded.”
It is important to note that when one talks about accessing public
sector information or open data, a single page on a data portal
such as data.gov.uk does not necessarily link to one item of data
or dataset – there can be multiple datasets.
Source: adapted from Data Protection Act and APPSI Glossary
De-anonymisation
The process of determining the identity of an individual to whom a
pseudonymised dataset relates.
Source: Open Data White Paper
Derived data
A data element adapted from other data elements using a
mathematical, logical or other type of transformation.
Source: OECD Glossary of Statistical Terms
Disclosive
Data is potentially disclosive if, despite, the removal of obvious
identifiers, characteristics of this dataset in isolation or in
conjunction with other datasets in the public domain might lead to
identification of the individual to whom a record belongs.
Source: Open Data White Paper
Executive agencies
A diverse group of organisations delivering a variety of services to
internal and external customers. They are part of the Crown and
do not usually have their own legal identity, but operate under
powers that are delegated from Ministers and Departments.
Source: OFT CUPI
Exhaust data
Data that is generated as a ‘by product’ of an organisation’s
activities, i.e. where data has been generated as a result of other
activities but not as the primary purpose of these activities.
Full cost pricing
A pricing policy in which charges are set to recover the full
resource costs of the activity.
Source: OFT CUPI
Free at point of use
Where there is no charge or fee to the end-user for the use or reuse of information.
Source: A Consultation on Data Policy for a Public Data
Corporation – glossary
Freemium
A business model by which a product or service is provided free
of charge, but a premium is charged for advanced features or
functionality.
147
Market Assessment of Public Sector Information
Source: APPSI Glossary
Geospatial data
Data or information that identifies the geographic location of
features and boundaries on Earth, such as natural or constructed
features, oceans and more.
Source: adapted from APPSI Glossary
Identifier
A particular element or reference in a dataset that allows
individuals’ or businesses’ identity to be known.
Information
Output of such process that summarises, interprets or otherwise
represents data to convey meaning.
Source: Open Data White Paper
Information asset register
Registers specifically set up to capture and organise metadata
about the vast quantities of information held by government
departments and agencies. A comprehensive IAR includes
databases, old sets of files, recent electronic files, collections of
statistics, research and so forth.
Source: OFT CUPI Open Data Handbook and Open Data Manual
Information fair trader
scheme
A scheme to set and assess standards for public sector bodies in
allowing the re-use of their information. Any public sector body
may apply to become IFTS accredited. However, all Crown
bodies that hold a delegation of authority from the Controller of
HMSO must become IFTS accredited. IFTS measures members'
performance against the six principles of maximisation, simplicity,
transparency, fairness, challenge and innovation. It considers
both the commercial re-use of public sector information and noncommercial citizen access to information.
Source: The National Archives Information Management Glossary
Intellectual property
A set of property rights that grant the right to protect the materials
created by them. Intellectual property comprises among other
things copyright, designs, patents, database rights, certain
confidential information and trademarks.
Source: Open Data White Paper and OFT CUPI
Licence
Permission by the copyright holder to reproduce or re-use
material protected by copyright.
Source: OFT CUPI
Linked data
The term used to describe the recommended best practice for
exposing, sharing and connecting items of data on the semantic
web using unique resource identifiers (URIs) and resource
description framework (RDF).
Source: APPSI Glossary
Market failure
An instance where a market is not efficiently allocating goods and
services. Market failures can include information asymmetries,
non-competitive markets, principal-agent issues and externalities /
public goods.
Marginal cost pricing
The cost of supplying another unit. Long run marginal cost is the
full extra cost (both fixed and variable) of providing a further unit
of output. Short run marginal cost measures how variable costs
change when output alters. In practice, marginal costs are difficult
to observe, and average variable costs are used as a substitute
148
Market Assessment of Public Sector Information
for the concept of marginal costs.
Source: OFT CUPI
Metadata
Data that describes or defines other data. Anything that users
need to know to make proper and correct use of the real data, in
terms of reading, processing, interpreting, analysing and
presenting the information. Thus metadata includes file
descriptions, codebooks, processing details, sample designs,
fieldwork reports, conceptual motivations, etc., in other words,
anything that might influence the way in which the information is
used.
Source: OCED Glossary of Statistical Terms and
www.sasc.co.uk/Guides/metadata.htm
Mosaic effect
The process of combining anonymised data with auxiliary data in
order to reconstruct identifiers linking data to the individual it
relates to. Also referred to as the ‘jigsaw effect’ or ‘crossreferencing’.
Source: adapted from Open Data White Paper
Non-commercial
government licence
A legal solution to enable the provision and use of public sector
information under a common set of terms and conditions at no
charge for non-commercial use only. The main requirement for reusers is to attribute the information provider and source.
Source: UK Government Licensing Framework
Non-departmental public
body
A body which has a role in the process of national government,
but is not a government departments or part of one, and therefore
operate to an extent at arm's length from Ministers.
Source: OFT CUPI
Open access
At its most narrow, this refers to the provision of free access to
peer-reviewed academic publications and other information data
to the general public.
Source: modified from Open Data White Paper
Open data
Data that meets the following criteria:
(a) accessible (ideally via the internet) at no more than the
cost of reproduction, without limitations based on user
identity or intent;
(b) in a digital, machine readable format for interoperation
with other data; and
(c) free of restriction on use or redistribution in its
licensing conditions.
Open data can be provided by the public and private sector as
well as individuals.
Source: adapted from Source: APPSI Glossary and Open Data
White Paper
Open government data
Public sector information that has been made available to the
public as Open Data.
Source: Open Data White Paper
Open Government Licence
The Open Government Licence (version 1.0), which forms part of
the UK Government Licensing Framework, offers a legal solution
to enable the provision and use of public sector information under
149
Market Assessment of Public Sector Information
a common set of terms and conditions. It enables any public
sector information holder to make their information available for
use and re-use under its terms.
Source: modified from UK Government Licensing Framework
Personal data
As defined by the Data Protection Act 1998, personal data means
“data which relate to a living individual who can be identified -
(a) from those data; or
(b) from those data and other information which is in the
possession of, or is likely to come into the possession
of, the data controller,
(c) and/or includes an expression of opinion about the
individual and any indication of the intentions of the
data controller or any other person in respect of the
individual.
Source: adapted from Open Data White Paper
Price elasticity of demand
A measure of the responsiveness of the quantity demanded of a
product or service following a change in its price.
Producer surplus
The value accruing to producers of goods and services when their
output is purchased by consumes. In traditional supply and
demand analyses, producer surplus is calculated as the
difference between the lowest amount the producer would be
willing to sell the good/service for and the price the producer
actually sold it for. In many cases this is equal to profit.
Processing
Under the Data Protection Act, processing in relation to
information or data, means obtaining, recording or holding the
information or data or carrying out any operation or set of
operations on the information or data, including –
(a) organisation, adaptation or alteration of the information
or data;
(b) retrieval, consultation or use of the information or data;
(c) disclosure of the information or data by transmission,
dissemination or otherwise making available; or
(d) alignment, combination, blocking, erasure or
destruction of the information or data.
Pseudonymised data
Data relating to a specific individual where the identifiers have
been replaced by artificial identifiers to prevent identification of the
individual.
Source: Open Data White Paper
Public domain
Works that are publicly available and in which the intellectual
property rights have expired or been waived.
Source: APPSI Glossary
Public good
A good or service provided by Government for public
consumption to combat an actual or perceived market failure.
Whilst these are generally free-at-the-point of use, for a good to
be public it has to be non-excludable and non-rivalrous. The
former meaning that no party can be excluded from the benefits
conveyed, whilst the latter means that one person’s consumption
does not prevent other persons from benefitting. PSI may be
150
Market Assessment of Public Sector Information
considered a public good in certain instances.
Public sector body
The State, regional or local authorities, bodies governed by public
law and associations formed by one or several such authorities or
one of several such bodies governed by public law. (Directive
2003/98/EC on the reuse of public sector information, Art 2).
Source: OFT CUPI
Public sector information
Public Sector Information covers the wide range of information
that public sector bodies collect, produce, reproduce and
disseminate in many areas of activity while accomplishing their
public tasks.
Source: adapted from BIS and APPSI Glossary
Public sector information
holder
A public sector body that collects and/or holds information, data or
content (as defined).
Source: OFT CUPI
Public task
Public task information is that which a public sector body must
produce, collect or provide to fulfil its core role and functions,
whether these duties are statutory in nature or are established
through custom and practice. The term 'public task' features in the
Re-use of Public Sector Information Regulations 2005 (SI 2005
No. 1515) and the INSPIRE Regulations 2009 (SI 2009 No.
3157). The National Archives provides guidance that assists
public sector bodies to define and publish a statement of their
respective public tasks.
Source: The National Archives
Raw data
In the context of public sector information, raw data is data
collected which has not been subjected to processing or any other
manipulation beyond that necessary for its first use. Raw data, i.e.
unprocessed data, is a relative term; data processing commonly
occurs by stages, and the 'processed data' from one stage may
be considered the 'raw data' of the next.
Source: A Consultation on Data Policy for a Public Data
Corporation – glossary
Refined data
This is where unrefined information has been enhanced,
manipulated and/or added to other inputs to create a retail
product for businesses or consumers. The process of refining
information can be undertaken by a PSIH, or viably in a
commercial market by the private sector.
Re-use
Use of information other than for the purpose it was originally
produced. This use could be for commercial or non-commercial
purposes. (Re-use of Public Sector Information Regulations 2005
(SI 2005/1515), Reg.4(2)).
Source: adapted from EU PSI Regulations
Resource description
framework
A W3C standard that is the foundation of several technologies for
modelling distributed knowledge and is meant to be used as the
basis of the Semantic Web.
Source: APPSI Glossary
Semantic web
A web of data that can be processed directly and indirectly by
machines.
151
Market Assessment of Public Sector Information
Source: APPSI Glossary
Sensitive personal data
This refers to personal data consisting of information as to a
person’s:
(a) racial or ethnic origins;
(b) political opinions;
(c) religious beliefs or other beliefs of a similar nature;
(d) whether they are a member of a trade union (within the
meaning of the Trade Union and Labour Relations
(Consolidation) Act 1992);
(e) physical or mental health condition;
(f) sexual life;
(g) the commission or alleged commission by them or any
offence; or
(h) any proceedings for any offence committed or alleged
to have been committed by them, the disposal of such
proceeding or the sentence of any court in such
proceedings.
Standard Industrial
Classification
First introduced in the UK in 1948, this is a framework for
classifying business establishments and other statistical units by
the type of economic activity in which they are engaged. There
are a number of levels of the classification, with subsequent levels
becoming more detailed.
Standard Occupational
Classification
A common classification framework of occupational information
for the UK on the basis of skill level and skill content.
Star Rating
In UK Linked Data, a system of ranking data sources that
indicates ease of machine readability. APPSI subjective score for
quality of a definition (qv).
Source: APPSI Glossary
Trading Fund
A government department, executive agency, or part of
department, established as a Trading Fund by a Trading Fund
Order made under the Government Trading Funds Act 1973. A
Trading Fund has authority to use its receipts to meet its
outgoings.
Source: adapted from HM Treasury Glossary (Managing Public
Money)
Unrefined data
This is data which cannot be substituted directly from other
sources. It relates to a PSIH’s monopoly activities, where
competition is very unlikely. Once a PSIH does something with
the data which could be performed viably in a commercial market
by the private sector it becomes refined information.
Value-added data
Raw data to which value has been added to enhance and
facilitate its use and effectiveness for the user. Value can be
added in a number of different ways including further
manipulation, compilation and summarisation into a more
convenient form for the end-user; editing and/or further analysis
and interpretation; and commentary beyond that required for
policy formulation by the relevant government department with
policy responsibility.
152
Market Assessment of Public Sector Information
Source: adapted from A Consultation on Data Policy for a Public
Data Corporation - glossary
Verbosity
The level of detail of a given dataset.
Velocity
How often data or information in a given dataset changes.
153
Appendix 2: Acronyms used in this
report
API
Application Programming Interface
APPSI
The Advisory Panel on Public Sector Information
BIS
Department for Business, Innovation & Skills
CUPI
Commercial Use of Public Information
DfE
Department for Education
DfT
Department for Transport
DSB
Data Strategy Board
EC
European Commission
EU
European Union
HMG
Her Majesty’s Government
HMT
Her Majesty’s Treasury
IAR
Information Asset Register
IFTS
Information Fair Trader Scheme
MoJ
Ministry of Justice
NDPB
Non-Departmental Public Bodies
NHS
National Health Service
ODI
Open Data Institute
ODUG
Open Data User Group
OFT
Office of Fair Trading
OGL
Open Government Licence
OS
Ordnance Survey
PSI
Public Sector Information
PSIH
Public Sector Information Holder
SIC
Standard Industrial Classification
Market Assessment of Public Sector Information
SOC
Standard Occupational Classification
SSNIP
Small but Significant and Non-transitory Increase in Price
TfL
Transport for London
155
Appendix 3: Literature Review
This Appendix surveys the literature that has analysed the role of public
sector information in economic growth, both in the UK and
internationally. Key themes are the direct economic value of public
sector information, the less tangible impacts on economy and society,
and debates over the most effective pricing models for public sector
information.
Overview
“A new market for public service information will thrive if data is freely available in a standardised
format for use and re-use...At present the market for information on public services is highly
underdeveloped. Open Data across government and public services would allow a market in
comparative analytics, information presentation and service improvement to flourish. This new
market will attract talented entrepreneurs and skilled employees, creating high value-added
services for citizens, communities, third sector organisations and public service providers,
developing auxiliary jobs and driving demand for skills.” 188
In so saying, the Government’s Making Open Data Real consultation laid out the argument for the
economic impact of increasingly the availability of public sector information for use and re-use. The
argument is that for public and private sectors alike, more widespread availability of public sector
information will drive efficiencies, boost innovation, lower barriers to entry and enable new insights
into old problems. In so doing it holds the potential to reduce public expenditure, promote
economic growth and make life more convenient for citizens. The objective of this Appendix is to
review the arguments and evidence around this view.
There is a considerable body of literature supporting the view that opening public sector
information to public use will boost innovation and economic growth. In this context, opening up
public sector information is often synonymous with making it available free of charge. However,
while the logic behind this argument is reasonably clear, the evidence to prove the point is often
incomplete. As discussed in the main report, this paucity of evidence arises for two principal
reasons.
The first is the complexity of measuring the impact of information, which permeates the economy in
so many ways. The direct revenue generated though sale of public sector information may be
measurable, where the data is available, but assessing the impact further downstream in the value
chain is much more complex and inevitably rests on a range of more or less defensible
assumptions. Where public sector information is released free of charge there is no direct revenue
to measure. The question becomes still more complex when factoring in possible substitutes for
public sector information.
The second reason is the difficulty of quantifying innovation. Part of the rationale for making public
sector information publically available is that this will in effect ‘crowd source’ innovations, as
thousands or even millions of users, it is hoped, experiment with new ways of using the data. This
process is expected to generate innovative new products and services that will benefit individual
users and society as a whole. While this has demonstrably taken place, and continues to occur, as
ever greater volumes of public sector information are made available to users, the process of
innovation is by its very nature unpredictable and, more often than not, there is a significant time
lag between data release and crystallisation of benefits. It is therefore difficult to quantify the
188
Making open data real: a public consultation, available at www.gov.uk/government/consultations/making-open-datareal.
Market Assessment of Public Sector Information
benefit, in terms of enhanced innovation, that is likely to arise from making any unit of public sector
information available to the public.
The purpose of this literature review is to provide an overview of the key literature on public sector
information, including:

attempts to size the economic contribution of public sector information;

assessments of the less tangible social, environmental and political value of public sector
information;

studies on the optimal pricing of public sector information; and

a series of case studies to illustrate these points.
The economic impact
That a link exists between public sector information and economic growth is widely accepted. A
recent review of the literature on public sector information (‘Review of recent studies on PSI reuse and related market developments’) concluded that “knowledge is a source of competitive
advantage in the “information economy”, and for this reason alone it is economically important that
public information is widely diffused,” listing benefits from public sector information including:
 development of new products built directly on public sector information;
 development of complementary products such as new software and services;
 reduction of transaction costs in accessing and using information;
 efficiency gains in the public sector itself; and
 the crossing of different public and private information to provide new goods and services 189 .
The economic importance of public sector information is seen to have increased radically with the
spread of new communication technologies, most notably the internet, and the development of a
‘knowledge economy’ in which value is generated through innovation in information and services.
These changes have had the effect of allowing the rapid diffusion of information to a large number
of end users, who not only benefit (individually and as communities) from the educational, political
and social advantages conferred by this information, but are also empowered to use it to create
innovative value-added goods and services. Furthermore, barriers to entry are reduced for startups when they enjoy free or cheap access to a wealth of data as well as the tools to easily reach
their target market. Although innovation cannot be manufactured by government, the conditions for
it to emerge and flourish can be created: this means that “enlarging and systematically inviting
serendipity can be argued to be an aim of government information policy making access to public
sector information an important cornerstone in a comprehensive digitally driven innovation
policy 190 .”
Conceptually, one can see how public sector information can drive economic growth and wider
prosperity. The simplified framework, shown below, illustrates the long-term drivers of economic
growth. Measurable outputs (such as GVA, employment and productivity) as well as less easily
measured outcomes (such as happiness and sustainability) are determined by a number of inputs
to the economy. In the long-term and considering the supply side of the economy only, the
economic output of the UK is a function of only two things: the number of people engaged in
gainful employment and the amount each person in employment is capable of producing.
189
190
Graham Vickery, “Review of recent studies on PSI re-use and related market developments” (2011)
Ibid.
157
Market Assessment of Public Sector Information
Figure A3.1: Long-term UK economic growth framework
Source: Deloitte analysis based on HMT and BIS analysis. See www.hm-treasury.gov.uk/d/ACF1FBD.pdf for more
details.
Drawing on the HMT’s research into the Five Drivers of Productivity 191 , the amount each worker
produces is determined by skill levels; the extent of innovation in products and processes; the
degree of investment in capital; entrepreneurial activity; and, lastly, levels of competition.
Underpinning employment and productivity are seven necessary enablers. These are related to the
infrastructure required to facilitate long-term economic growth. A deficit in these enablers will
equate to a supply-side constraint on economic growth in the long-run. This could be caused either
by limiting the growth in working population (through insufficient housing capacity or supporting
utilities) or by acting as a drag on productivity growth (through below-par ICT connectivity or a
sclerotic transport system). Expanding and enhancing these enablers can therefore positively
impact employment and productivity, which in turn generate economic growth and greater
prosperity.
As Figure A3.1 shows, improvements in the availability and quality of public sector information can
impact across all infrastructure enablers and productivity drivers – the following sections explain
how.
The economic contribution of public sector information
Despite this consensus around the importance of public sector information in a modern ‘knowledge
economy’, however, there is little agreement when it comes to quantifying the economic benefit
contribution of public sector information. Attempts have varied widely both in the methodologies
employed and in the conclusions reached, with the value added by public sector information
assessed at figures ranging from millions to hundreds of billions of pounds. This is unsurprising
given the difficulties mentioned in the introduction to this chapter: the ubiquitous presence of public
191
See www.bis.gov.uk/analysis/economics/productivity-and-competitiveness for further details. 158
Market Assessment of Public Sector Information
sector information, making its impact difficult to disentangle from other factors; and the challenges
in predicting and quantifying innovation.
Several recent studies have attempted to tackle this challenge. They have examined three levels of
economic value generated by public sector information:

direct value: i.e. revenue generated by government from selling access to public sector
information;

commercial value: i.e. the revenue generated by private companies through the use of public
sector information; and

downstream value: i.e. the value to users of products and the wider economic, social and
environment benefits generated.
Assessing direct value is the most straightforward. A 2006 study by the Office of Fair Trading
(OFT) (‘The Commercial Use of Public Information’) estimated the direct revenues from around
400 UK PSI holders at £400m. The report extended this analysis by assessing the combined
consumer surplus (the summed difference between the highest prices consumers would pay and
the actual price) and the producer surplus (the difference between the price of the product and the
cost of supplying it). This led to an assessment of economic value of £590m. The OFT estimated
the total potential value of public sector information in the UK to be roughly double this figure, at
£1.11 billion if three kinds of market distortions to be removed: unduly high pricing, distortion of
downstream competition, and failure to exploit public sector information.
The OFT’s approach avoids the danger, common with the top-down approach, of over-estimating
value. However, it is likely that the results of its analysis significantly understate the value, both
actual and potential, of public sector information. This is partially because some public sector
information providers were outside the scope of the report, so that significant volumes of public
sector information are excluded from the analysis. More fundamentally, the report did not consider
the wider uses and re-uses of public sector information and the potential for enhancing the growth
of existing companies and triggering new innovations and entrants into the market. By way of
contrast, Vickery’s 2011 literature review on reusing public sector information valued public sector
information across the entire EU at EUR 140 billion. 192 He derived this estimate by extrapolating
figures for the total economic impact of geospatial information in Australia and New Zealand.
Based on this calculation, the Government in its Autumn Statement 2011 estimated the current
total economic value of public sector information in the UK at £16 billion. 193 The significant
difference between Vickery’s estimate and the OFT’s figures demonstrates the impact of adopting
a top-down approach to assessing value.
Other studies that have attempted to assess the commercial value of public sector information
have on occasion reached even higher figures. The most relevant UK studies that have attempted
to put a value on public sector information include:

a 1999 Oxera study (‘The Economic Contribution of Ordnance Survey GB’) which
concluded that £79-£136 billion of Gross Value Added was dependent to some extent on
products and services provided by Ordnance Survey. This study employed the value-added
method, calculating the total value of all the products and services produced in the UK in which
Ordnance Survey’s products and services serve as an input.

a 2007 study by PA Consulting (‘The Public Weather Service’s Contribution to the UK
Economy’) which concluded that the public valued the services of the Met Office at £353.2
million, with a minimum additional contribution to the UK economy of £260.5 million based on
three case studies.
192
193
Vickery (2011)
NAO, Implementing Transparency (2011)
159
Market Assessment of Public Sector Information

a 2003 study on the British Geological Survey (‘The Economic Benefits of the BGS’) which
estimated its contribution at between £34bn and £61bn, again using the value-added method.
An additional study of interest was carried out in 2010 by ConsultingWhere and ACIL Tasman for
the Local Government Association (‘The Value of Geospatial Information to Local Public
Service Delivery in England and Wales’). The authors calculated that the benefit to the local
government sector from the use of geospatial data was £232 million over five years, with a GDP
increase of £323 million for England and Wales (equivalent to 0.02 per cent of GDP for England
and Wales) and increased taxation of £44 million. The report also calculated that geospatial data
had led to an increase in labour productivity of 0.233 per cent among local public sector providers,
equivalent to an additional 1,500 full time staff in England and Wales. 194 Since this relates only to
one type of data in one sector, it is to be supposed that the overall effect of public sector
information would be calculated to be much higher were the same methodology applied to all
public sector information across the whole economy.
Figure A3.2 summarises the results, method and scope of these studies.
Figure A3.2: summary of public sector information market value in literature
Notes:
‐ This diagram is indicative rather than scientific – it is intended to comparatively illustrate the differing methodological approaches and
conclusions of the reports discussed in this chapter
‐ In general the split is between reports which examine the impact of a specific provider of PSI, for example the Ordnance Survey, and
reports which attempt to assess the impact of all PSI across the whole economy. In addition, there is a division between reports which
adopt a top-down approach and those (notably the OFT report) which adopt a bottom-up approach
‐ The figures shown here should therefore not be treated as directly comparable – the OFT report was not attempting to measure the
same things as, for example the Pira International report and this accounts for some of the difference in their conclusions
In 2000 a study was undertaken by Pira International (‘Commercial Exploitation of Europe’s
Public Sector Information’) which estimated the national income attributable to economic
activities based on the exploitation of public sector information at EUR 68 billion, or 1.4 per cent of
EU GDP. Costs, or the government investment in collecting public sector information, were valued
at EUR 9.5 billion, equating to a seven-fold return on investment. The UK’s share of this value was
estimated to be EUR 11.2 billion (with an upper estimate of EUR 21.8 billion and a lower estimate
194
ConsultingWhere and ACIL Tasman, ‘The Value of Geospatial Information to Local Public Service Delivery In England
and Wales’ (July 2010)
160
Market Assessment of Public Sector Information
of EUR 4 billon). These estimates were based not only on the direct revenues from the supply of
raw public sector information, but also on value generated by products developed using public
sector information as an input. While these Pira values are more moderate, and therefore perhaps
more defensible than some of the figures reached by the studies mentioned above, the
considerable range of the estimates produced by the report indicates that this is not a precise
science. 195
The wide variation in estimated value results from differing methodologies, assumptions and
economic models. The OFT report was critical of top-down assessments, arguing that this
approach is prone to overstating the value of public sector information since it fails to account for
potential substitutes. For example, it is clearly not true that without the information provided by
Ordnance Survey there would be no mapping information available to users (a point Oxera
acknowledge in their own report). Commercial customers have at their disposal a range of potential
suppliers of this information – although it may be true that Ordnance Survey provides information
that is more comprehensive and of a higher quality than would be possible for a private sector
provider, given the investment required. It is therefore not reasonable to claim that the all the value
added to the economy by Ordnance Survey information would not be created were Ordnance
Survey not to exist. It is also clearly problematic drawing parallels between countries, given
variations in the supply and demand for public sector information as well as associated costs,
licensing conditions, and other differences. Even drawing parallels between different types of data
is fraught with difficulties, as it is clear that some types of data have generated more value than
others. Nonetheless, as noted, the OFT’s approach is also not without its drawbacks.
The costs of providing public sector information
As the NAO concludes in a recent report on the transparency agenda (‘Implementing
Transparency’), “when estimates of economic value vary this widely, it is difficult to assess the
scale of effort or targeting needed to best build on that value.” 196 This uncertainty matters because
the cost of making public sector information available to the public must be offset against the
forecast value created in order to create a robust business case. Unfortunately, just as there is
very little certainty regarding the economic value of public sector information, there is also very little
data available on the costs of public sector information provision, at least beyond the major trading
funds and ONS. The NAO estimates additional staff costs of providing disclosure for pre-existing
data range from £53,000 to £500,000 annually by department. 197 However, this does not take into
account other costs, including those incurred by IT and other support functions, which are difficult
to disaggregate.
Examples of costs associated with public sector information made publically available to date also
vary. For example, the police crime map incurred set up costs of £300,000 and annual running
costs of over £150,000, largely because the department repackaged the information to improve
accessibility. 198 On the other hand, other cases such as the release of public weather service data
have incurred much lower costs. This suggests that, as might be expected, improving the quality of
data and providing an integrated platform for users is likely to require much greater investment of
resources than simply releasing the data in its existing form. However, in value for money terms,
this may under some circumstances prove more cost effective than releasing data which the
majority of non-technical users are unable to exploit.
Providing value for money from public sector information
195
It is also worth noting that the report concluded that commercial activities based on PSI have considerable scope for
expansion in the EU. AS support for this view, it noted that value generated from the use of PSI in the United States is
between two and five times greater than in the EU.
196
NAO, Implementing Transparency (2011)
197
NAO (2011)
198
NAO (2011)
161
Market Assessment of Public Sector Information
The investment into providing public sector information may be judged to provide value for money
if improved quality means the data enjoys improved usage and therefore provides greater
economic and social benefits. It is notable that releases of data have attracted varying levels of
interest. For example, the police crime map website received around 47 million visits between
February and December 2011. In contrast, the NAO reports that over 80 per cent of visitors to
data.gov.uk leave without clicking on any links, and departments report limited interest in
departmental spend data releases. 199 This disparity may be a consequence of the police crime
map presenting data in a format that appears much more user-friendly than data.gov.uk for the
average user, commensurate with the significant investment made. Similarly, government
spending data has been released in a relatively unprocessed format and may therefore be
challenging for the majority of users to exploit. In support of this view, the NAO records that “the
Department for Education has reported an 84 per cent increase in the use of its comparative data
on schools, compared with the same period last year, since it was consolidated in one location and
data were made more accessible 200 .”
It therefore seems apparent that while making public sector information available is an important
first step, it is unlikely to enjoy widespread use by the public unless it is presented in a user-friendly
format, consolidated in one easy to find location: data availability is a necessary, but not always
sufficient condition for use and value creation, It may be that the better prepared the data, the more
likely it is to produce significant economic and social value as a wider range users are able to
deploy it to improve accountability and develop innovative products. This requires greater
investment of resources by public sector bodies responsible for publishing the data, but it is a
reasonable hypothesis that, at least for some types of data, the investment will be rewarded by
strong economic and social returns. As the NAO argues: “evidence on benefits should be
considered alongside information on costs and risks to secure best value from the large stock of
public data, match the range and presentation of data purposefully to fulfil specific objectives,
ensure that risks are identified and mitigated and secure value for money 201 .” However, in order to
make the business case for investment, and ensure that this investment is appropriately targeted
towards the most valuable PSI, it would be helpful to have a more comprehensive understanding of
both the benefits and the costs.
Summary
There appears to be little dispute that public sector information generally delivers economic and
social value in excess of the costs required to provide access to it. However, accurately assessing
its economic contribution, both actual and potential, requires an improved understanding of the
costs and benefits associated with public sector information, including:

a better understanding of drivers of additional costs to release different types of public sector
information;

distinguishing between producer surplus and consumer surplus benefits from releasing more
public sector information;

a clearer means of determining demand, so as to prioritise the release of public sector
information; and

a robust method of evaluating the emerging effects of public sector information as it is
released, so that efforts can potentially be focussed on high-value data.
It is important to note that all actors could deliver more in terms of transparency of information on
the topic. Public sector bodies have, in some instances, not responded to survey questions posed
199
It should be noted that data.gov.uk has been developed since 2011 and the situation may have developed since the
NAO published their report
200
NAO (2011)
201
NAO (2011)
162
Market Assessment of Public Sector Information
as part of the study, and naturally, private sector organisations are unwilling to discuss and
disclose sensitive financial information and the precise means by which PSI does or could improve
performance.
Figure A3.3 below summarises the key studies that have attempted to value the contribution of
public sector information. It should be noted that although summarised here for convenience, these
reports were not all attempting to measure the same thing and so the figures are not directly
comparable.
Figure A3.3: summary of UK public sector information studies
Study
Date
Author
PSI sources
evaluated
Upper value
estimated
Lower value
estimated
OFT CUPI
2006
OFT
Various
£1.1bn
(potential)
£590m
Vickery
2011
Graham Vickery
Various
c. £16bn (UK);
EUR 140bn
(EU)
-
MEPSIR
2006
HELM Group
Various
EUR 48bn (EU +
Norway)
EUR 10bn (EU
+ Norway)
PIRA
2000
Pira International
Various
EUR 21.8bn
EUR 4bn
The Economic
Contribution of
Ordnance
Survey GB
1999
Oxera
Ordnance
Survey
£136bn*
£79bn*
The Public
Weather
Service’s
Contribution to
the UK Economy
2007
PA Consulting
Met Office
£353m
-
The Economic
Benefits of the
BGS
2003
Roger Tym &
Partners
British
Geological
Survey
£61bn*
£34bn*
Source: Deloitte analysis
* These figures are based on an analysis of the Gross Value Added dependent to some extent on PSI, and therefore differ from
attempts to assess the value of PSI
The broader impact
Many of the studies discussed above attempt to quantify the actual or potential economic
contribution of public sector information. In reality this is difficult to separate from other impacts,
such as the social, environmental or political benefits conferred by access to public sector
information. For example, if an individual’s or a community’s health outcomes are improved
through use of public sector information , this will have a downstream economic benefit in terms of
increased individual productive potential (through avoidance of loss of working hours from illness)
and medical costs avoided. Due to the complexity of calculating these less tangible benefits,
however, they are often omitted from quantitative studies of economic contribution.
However, it is also true that these benefits have an intrinsic value – good health is desirable
irrespective of its positive economic impacts, and would remain so even if quantitative economic
impacts could not be demonstrated. For this reason it is important to take such benefits into
account when weighing up the costs and benefits of public sector information. Where it is not
possible to calculate a financial value, case studies and other measures can be used.
163
Market Assessment of Public Sector Information
One of the most important studies to address this area is the 2007 independent review by Ed Mayo
and Tom Steinberg (‘The Power of Information’), which cited a range of benefits that have
resulted from sharing information 202 :
Figure A3.4: broad benefits of public sector information
Information use
Benefits
Online communities
In medical studies of breast cancer and HIV patients, participants in online
communities understand their condition better and show a greater ability to cope. In
the case of HIV, there are also lower treatment costs.
This creates both a welfare benefit and a cost saving benefit.
‘Wired’ local
communities
Studies of ‘wired’ local communities demonstrate that there are more neighbours who
know the names of other people on their street.
This is likely to create a safer, more cohesive and supportive community.
Restaurant food safety
information
Sharing restaurants’ food safety information in Los Angeles led to a drop in foodborne illness of 13.3 per cent, compared to a 3.2 per cent increase in the wider state
in the same time frame.
The proportion of restaurants receiving ‘good’ scores more than doubled, with sales
rising by 5.7 per cent.
This creates a welfare benefit, reduced medical costs and increases consumer
spending.
Medical prescription
information
By providing clear information when dispensing medication, pharmacists can improve
patient adherence/persistence with medication advice by 16–33 per cent.
Source: Mayo and Steinberg, Deloitte analysis
Not all of these examples concern public sector information but they do indicate the range of
benefits which can be delivered by widespread availability of information. They therefore hint at the
potential benefits of making public sector information available to a wider range of users, given the
volume and range of information held by public sector bodies. It is important to ensure that these
less tangible and less easily quantifiable benefits are not excluded from any analysis of the costs
and benefits of public sector information.
In a more recent study, Pollock (2011) estimated that the indirect benefits from greater availability
of public sector information, such as reductions to transaction costs to users and re-users and
other efficiency gains could imply gains of around £600m per year for the UK.
Maximising the impact of public sector information
Pricing models
An important aspect of the debate over the economic contribution and value of public sector
information is the question of the correct level of pricing for data. Currently, while much public
sector information is made available at no charge for use and re-use, some of the most valuable
data – including, for example, some mapping data collected by Ordnance Survey, and data held by
Companies House – is only available to paying users. There is some debate over whether these
charging models are justified, given the costs involved in collecting the data, or whether this
introduces counterproductive inefficiencies and market distortions.
202
Mayo and Steinberg, ‘The Power of Information’ (2007)
164
Market Assessment of Public Sector Information
In one of the key papers on this topic (‘Models of Public Sector Information Provision via
Trading Funds’), Newbery et al. identified four possible charging policies:

Profit-maximization: setting a price that maximises profit given the demand for the data;

Average cost (cost recovery): setting a price equal to average long-run costs (including
fixed costs incurred by data production);

Marginal cost: setting a price equal to the actual cost of the process of supplying data to a
user; and

Zero cost: releasing the data free of charge. 203
The paper concluded that many Trading Funds’ products – primarily ‘refined’ data products – could
not be analysed effectively due to data limitations, and that these should therefore by default be left
with their pricing policies unchanged. However, Newbery t al argued that most ‘unrefined’ data
should be made available at marginal cost, which owing to the low cost of providing digital data
would effectively be zero. This would bring trading fund data into line with raw data charging
policies in other areas of government. The basis for this argument is that the benefits to society of
making this data freely available would outweigh the costs, although some trading funds, notably
Ordnance Survey, would need additional taxpayer support.
The costs and benefits estimated by the paper are summarised below:
Figure A3.5: costs and benefits of Trading Funds and other PSIHs
Trading Fund
Gross benefit from
releasing raw PSI
Cost to government
Net benefit
Companies House
£2.6m
£681k
£1.9m
The Met Office
£1.2m
£260k
£1.03m
Ordnance Survey
£168m
£12m
£156m
UK Hydrographic Office
£1.08m
£744k
£338k
The Land Registry
£2.3m
£1.1m
£1.2m
The DVLA
£4.3m
£582k
£3.7m
Source: Newbery et al, ‘Models of Public Sector Information Provision via Trading Funds’ (2008)
Clearly, this study raises important questions regarding the fairest way of financing trading fund
data collection activities: should taxpayers in effect subsidise free access to data for users, or
should users of data directly bear the costs associated with data collection and dissemination?
Newbery et al’s figures indicate that free access to trading fund data, subsidised by the taxpayer,
would in most cases lead to a net economic benefit. There is no guarantee, however, that this
benefit will be equally spread across society. It is possible, for example, that most of the benefit
might accrue to foreign-based companies which can use the data to boost revenues but not
contribute to UK tax revenues. If it could be demonstrated that making the data freely available
would be ‘revenue neutral’ – i.e. the associated tax income to government would at least equal the
outlay in support of trading funds – there would be a much more economically and politically
powerful case for taking this step.
In a 2009 paper (‘Enhancing access to government information: economic theory as it
applies to Statistics Canada’) Kirsti Nilsen argued that in the case of public sector information the
benefits are non-rivalrous and non-excludable, and so the principle that the beneficiary pays is a
fallacy.
“The justification for cost recovery is often based on the so-called benefit principle:
Those who benefit from a good should pay for it. However, it is very difficult to
203
Newbery et al, ‘Models of Public Sector Information Provision via Trading Funds’ (2008)
165
Market Assessment of Public Sector Information
determine the benefits of information. Information flows. It moves away from the
initial buyer. So, what is the benefit? Who benefits? How do you apply the benefit
principle? The assignment of benefit, like the assignment of costs, is an arbitrary
exercise.” 204
Nilsen argued that should Statistics Canada move from a cost recovery pricing model to a zero
cost model, the following results could be expected:

sales and licensing revenues for the public sector body would decrease;

usage and reuse of the public sector information would increase;

increased public sector information usage would provide positive externalities:
information dissemination, more widespread usage of the data, and economic growth;

tax revenues would therefore increase; and

the public sector body’s transaction and opportunity costs would decrease. This should
offset the decline in sales and licensing revenues, in addition to the increase in tax
revenues.
In considering these arguments, however, it is important to reflect on the diversity of the public
sector information landscape and the dangers of applying a ‘one size fits all’ prescription. For
example, the Trading Funds operate within a regulatory framework which limits their ability to
cross-subsidise information provision from income generated in other areas of activity. From
conversations with stakeholders it is clear that the picture is more complex than an initial
examination might suggest, as are the implications of radical changes to the current model.
Examples where this approach might be less appropriate are considered elsewhere in this report.
The impact of zero or marginal cost pricing on innovation
There is already a trend towards making at least some data freely available. This is clear across
the full spectrum of PSIHs, from government departments through to trading funds such as
Ordnance Survey, which now offers significant numbers of datasets through its Open Data
initiative. A 2011 European Commission study (‘Pricing of Public Sector Information’) found “a
clear trend towards lowering charges and/or facilitating re-use” 205 among public sector bodies. The
study found that where PSBs moved to marginal or zero cost charging, the number of re-users
increased by between 1,000 per cent and 10,000 per cent, indicating the potentially significant
impact of removing cost barriers to re-use. Removing cost barriers was also found to attract new
types of users, particularly SMEs, which may be a particularly innovative class of user and thus
add additional value to the economy. The study further found that costs did not increase
significantly, or even decreased, once prices were lowered. This is partly due to the fact that zero
cost pricing greatly reduces transaction costs, as public sector information providers no longer
need to devote resources to complex payment, licensing, supply and enforcement systems. In
addition, enhanced volumes of users compensate for reduced prices.
The study also found that the availability of low cost data with clear re-use rules spurred the
development of innovative public sector information-based apps, opening up a potentially
significant new market. It cites examples such as MetroParis and London Tube apps which have
jointly generated EUR 400,000 in revenues. However, when public sector information providers
create value-added services in the form of their own apps this has been noted to have a
detrimental effect on private sector app innovation, suggesting that they should arguably confine
themselves to providing data for users and avoid competing directly through the development of
value-added products. Some stakeholders, however, have made the point that private sector
204
205
Nilsen, Enhancing access to government information: economic theory as it applies to Statistics Canada (2009)
Pricing of Public Sector Information (2011)
166
Market Assessment of Public Sector Information
providers did not develop apps and other value-added products taking advantage of the available
data, prompting them to release their own to fill a gap in the market.
A significant body of work therefore supports the argument that the gross benefits from not
charging, or charging at marginal cost, outweigh the added cost that may be borne by government.
This has contributed to growing momentum towards reducing the cost of public sector information
at the point of use. Despite the strength of Nilsen’s arguments from the standpoint of economic
theory, however, there may be political challenges in supporting data collection which is then freely
provided to for-profit businesses, even if the wider benefits to society can be demonstrated to
outweigh the direct cost to taxpayers. The challenge for advocates of open data, therefore, is to
demonstrate that changes to pricing policy can be revenue neutral or lead to an increase in public
sector revenues, through a growing tax base.
There is evidence indicating that changes to pricing policy can be revenue neutral. A 2011 Finnish
study (‘Does marginal cost pricing of public sector information spur firm growth?’) surveyed
the performance of 14,000 firms in the architecture and engineering sector (Standard Industrial
Classification 7420) and pricing policies for public sector geographical information across 15
countries. It concluded that in those countries where geographical information was either free or
priced at marginal costs, firms grew 15 per cent faster than in countries where information was
priced at cost recovery. The study also found that an impact on company growth rates could be
detected within a year of switching to a marginal cost pricing scheme, although the impact became
more pronounced after two years. 206
Importantly, the most significant impact on growth rates was experienced by SMEs rather than
large firms. The author attributes this to high public sector information prices creating a barrier to
entry for SMEs, which is removed by a switch to marginal cost pricing. He argues that for this
reason, marginal cost pricing of public sector information is likely to create more competitive
markets and thus lead to lower prices and a wider range of products, benefitting consumers. If this
result could be demonstrated to be replicated across other types of public sector information, it
would present a persuasive argument – from the point of view of encouraging dynamic markets
and economic growth – for moving to a marginal cost pricing mechanism for public sector
information. Nonetheless, such a change would need to be evaluated on a case by case basis,
given the various types of information involved and the various markets in which public sector
bodies which currently charge for information operate.
The table below summarises the key studies that have examined the pricing of public sector
information.
Figure A3.6: pricing of public sector information studies summary
Study
Date
Author
Area examined
Conclusions
Models of Public
Sector Information
Provision via
Trading Funds
2008
Newbery et al
UK Trading Funds
Significant net benefit from
moving to marginal or zero cost
pricing structure for raw data
Enhancing access
to government
information:
economic theory
as it applies to
Statistics Canada
2009
Kirsti Nilsen
Statistics Canada
Zero cost pricing produces
positive externalities across
society while being revenue
neutral due to increased usage
and reduced transaction costs
Pricing of Public
Sector Information
2011
Deloitte and
others in
association with
Charging models for
PSI
Moving to marginal or zero cost
pricing increases users by
1000-10,000 per cent, attracts
206
Heli Koski, Does marginal cost pricing of public sector information spur firm growth? (2011)
167
Market Assessment of Public Sector Information
Study
Date
Author
Area examined
the European
Commission
Does marginal
cost pricing of
public sector
information spur
firm growth?
2011
Heli Koski
Conclusions
new types of users including
SMEs, and does not result in
significantly higher costs
Performance of
14,000 firms across
15 countries
Marginal cost pricing of
geographical data leads to an
average 15 per cent increase in
the growth rate of SMEs
Case studies
The economic and social value potentially generated by public sector information is well illustrated
by the case of ‘wicked’ problems. These are challenges featuring complex interdependencies,
which spill across jurisdictions and therefore have the potential to confound traditional, vertically
organised structures such as government departments, whose expertise and mandate covers only
one area of policy. Examples of wicked problems include climate change, youth unemployment
and health issues such as obesity. By making public sector information available for re-use data
from diverse sources can be linked and ‘mashed’, potentially producing new insights and
generating unexpected solutions.
The value that can be generated by public sector information is also well illustrated through case
studies on the impact on the public sector, business and wider society. These have been drawn
from the UK and other countries that have been experimenting with different models for providing
public sector information. While to some extent this is an anecdotal approach – it is not possible to
forecast the number of new businesses or their value – it offers useful insights into the value that
can be generated through public sector information.
Public sector efficiencies
Case study
Spotlightonspend
Helping public
sector bodies
publish data in a
cost-effective
manner
Barnet StreetPatrol
Delivering efficiency
savings in local
policing
Description
Spotlightonspend is a managed service that is comprised of everything necessary to
facilitate cost-effective publication of the spend and related information that is made
available to the public.
Spotlightonspend is designed to:

Cut costs by removing the need to add to the workload of current staff or increase
headcount

Eliminate the complexity of becoming and staying compliant with policy

Reduce the risk of inadvertent breach of data protection legislation

Enhance the information published to improve its accessibility, relevance and value
for the public
Barnet, one of the largest boroughs in London, deployed a GPS system called
StreetPatrol to help street wardens locate, identify and photograph issues including
abandoned vehicles, graffiti, antisocial behaviour and fly-tipping. This enables them to
send information immediately back to head office, ensuring a rapid and efficient
response. Previously this process could take between three and four days.
Using StreetPatrol, wardens are able to spend up to 70 per cent of their time on patrol,
compared to 30 per cent for those without the system.
Adoption of this system has delivered around £180,000 in efficiency savings. Improved
speed of response also brings cost savings: an abandoned vehicle costs
approximately £50 to recover; a burning vehicle nearly £4,000. By responding quickly
wardens are able to tackle problems before they develop.
Source: Deloitte: Open Data – driving growth, ingenuity and innovation; and Cabinet Office material
168
Market Assessment of Public Sector Information
Business innovation
Case study
Description
Red Spotted Hanky
Using rail industry data
to offer customers lowprice tickets
Launched in 2010, Red Spotted Hanky is an online ticket retailer which aims to offer
customers and easier way to book without any administration or payment fees.
Red Spotted Hanky relies on data from the rail industry to offer customers low-cost
advance bookings. The business employs 13 people and is growing fast, with a
loyalty scheme, a tie-up with Tesco and a ‘price promise’ for customers.
Duedil
Linking and
aggregating
information to improve
business transparency
Duedil is a business information provider based in London’s Soho. Duedil gives free
access to governance and financial information for every company in the UK and
Ireland, and combines this with data from online sources, Application Programming
Interfaces, social networks and more. Duedil was launched in April 2011 by
entrepreneur Damian Kimmelman. By late 2011 it as valued at £20 million and had
attracted considerable interest from investors.
Its aim is to make business more transparent, by opening up company information
to make the due diligence and research process simple and intuitive. By
aggregating and linking all the available information, users can gain a
comprehensive understanding of businesses and the people who run them.
Parkopedia
Using local authority
data to give drivers live
parking updates
Parkopedia is an innovative open data company which fuses location and other
data. A small UK-based business, it uses live data from local authorities to help
drivers identify free car parking spaces. Parkopedia has grown to become the
world’s leading source of parking information covering more than 20 million spaces
in 25 countries.
Used by millions of drivers, Parkopedia’s service include a pre-booking tool which
allows drivers to book parking online, and real-time parking space availability
information. Parkopedia also works with other organisations to integrate its data into
journey planner mobile applications and satnavs.
Source: Deloitte: Open Data – driving growth, ingenuity and innovation
Case study – the value of PSI in the mobile app market
One area where public sector information has generated significant innovation, some of which has been
successfully monetised, is in the mobile app market. Deloitte’s POPSIS study (2011) estimated the value of the
mobile app market at around $35 billion, of which 40% is contributed by apps that use public sector information.
Based on a UK contribution of 13% of all apps, this suggests that the value of the PSI-based app market in the
UK is some £1.13bn.
Furthermore, this is a rapidly growing market, with use of smartphones and tablets in the UK increasing rapidly
year on year. The figure of £1.13bn is likely to understate the true value of public sector information-based apps
to the UK economy, as it does not take account of the benefit to users (such as time savings and increased
ability to access important information).
Country distribution of mobile apps in the sample
Public sector information-based app share
169
Market Assessment of Public Sector Information
Based on ‘Apps’ market snapshot of POPSIS study. The study considers a sample of ~500 apps from Android,
Istore, and Ovi apps
Source: Deloitte, available http://ec.europa.eu/information_society/policy/psi/docs/pdfs/report/11_2012/apps_market.pdf
Country case studies
While international case studies are a helpful illustration of public sector information practices
elsewhere, care must be taken when extrapolating the lessons. For example, the same proportion
of benefits may not occur in the UK due to differences in economic structures, legislative
frameworks and the culture and capabilities of consumers.
Denmark’s Open Data Innovation Strategy 207
In 2009 the Danish government launched the Open Data Innovation Strategy (ODIS) to provide
easier access to PSI for businesses and other users. Although this is a recent development that is
likely to have much as yet unrealised potential, efficiency gains in both the public and private
sectors are already emerging. Certain sectors have been particularly quick to spot the
opportunities presented by the availability of public sector information.

Financial services: banks are working with the tax authorities to access payroll and
pension data for clients of the banks. This has the estimated potential to save banks EUR
67 million per year in efficiency gains and reduced losses. Additional savings could be
gained by accessing data on clients’ employment conditions. The insurance industry is also
interested in the potential for customer data to improve its risk assessments. In these
cases, however, there remain issues to be resolved around customer consent for the
sharing of these personal details.

The energy sector: data on building specifications, as well as demographic data on
residents, could be used to target energy-saving measures. Potential savings are estimated
at EUR 0.54-2.7 billion.

The pharmaceutical and healthcare sector: patient data could be used to select patients
for clinical trials, improving the process of developing new drugs.
Denmark is in some ways an unusual case owing to its relatively small size and a legacy of robust
public sector data collection and digitisation, which has left it well placed to improve economic and
social outcomes through the release of public sector information. It is also clearly at an early stage
in this process, with significant issues in data security and confidentiality yet to be resolved.
Nonetheless it provides a striking example of the potential benefits to be realised by making public
sector information available to business users, both in terms on financial savings and wider
benefits to society.
Denmark: the value of address data 208
In 2002, the official Danish address data was made available free of charge. This meant that any
user could access the data without paying a fee to the Danish Enterprise and Construction
Authority (DECA) charged with collecting and maintaining this data. Eight years later DECA
commissioned a study to analyse the benefits of making the data free of charge.
The study concluded that between 2005 and 2009 the total direct financial benefits of the data
were EUR 62 million, with costs of around EUR 2 million. The study also estimated that in 2010
total benefits would be EUR 14 million, with costs of EUR 0.2 million. The benefits were split at
around 30 per cent in the public sector and 70 per cent in the private sector.
207
208
Vickery (2011)
‘The value of Danish address data’ (2010)
170
Market Assessment of Public Sector Information
As with many other studies, only direct financial benefits were measured, as downstream
economic and wider social benefits were considered too difficult to quantify. However, the authors
suggested that these benefits were likely to be of considerable value. They cite the example that
46 per cent of Danish families own a GPS navigation system, incorporating the address datasets,
which has doubtless generated significant benefits – for example in travel time saved.
The study also highlighted barriers which have limited the benefits of the address data, caused by
technical, traditional and legislative barriers which, for example, limit how quickly business
addresses are registered and included in the database. This indicates the need for efficient data
collection, processing and dissemination systems and the regulatory framework to support this.
In conversations with stakeholders, the caveat has been raised that some of the benefits identified
in this study may not be transferable to the UK. In particular, the Public Sector Mapping Agreement
means that sharing of geospatial data between public sector bodies in England and Wales 209 is
already relatively advanced compared to many countries. Utilities companies also have access to
this information. Furthermore, the UK has an unusually dense post code network, which means
that these are suitable for accurate navigation via GPS navigation systems. These caveats
highlight the risks inherent in assuming that benefits experienced in one country will be directly
transferable to another.
The study nonetheless highlights the important direct financial benefits that can result from making
these key datasets freely available to users, and suggests that these direct financial benefits both
in the public and private sector can considerably outweigh the cost to government.
Spain: The Aporta Project
In 2011 an analysis of the ‘infomediary’ business sector was carried out, defined as “the set of
companies that create applications, products and/or added-value services for third parties, using
public sector information.” 210 This was intended to be a comprehensive survey of all Spanish PSI
activities. Headline findings were:

business turnover directly associated with infomediary activities is EUR 550-650 million, 35-40
per cent of the total company activity of EUR 1.6 billion;

activity by re-use field was as follows:
o
business/financial 37.6 per cent
o
geographic/cartographic 30.5 per cent
o
legal 17 per cent
o
transport 5.2 per cent
o
social data/statistics 1.9 per cent
o
meteorological 1.1 per cent
o
others 6.7per cent

the re-used information comes mostly from national agencies, but half of the companies also
re-use international information;

re-use policies are valued, particularly to improve the quality and accuracy of information,
improve understanding of the legal framework, and expand the amount and scope of
information generated; and

areas identified for improvement include standardization of formats, standardization and
improvement in the regulation of licences for re-use, and pricing of information.
209
210
The One Scotland Mapping Agreement covers public sector bodies in Scotland.
Vickery (2011)
171
Market Assessment of Public Sector Information
This survey indicates the scale of the opportunity for business stemming from the re-use of public
sector information, with the financial services and geospatial sectors seeing particularly significant
opportunities. However, it also highlights the need to improve quality and accessibility, through
standardised formats and improved licensing arrangements.
Australia: Office of Spatial Data Management and Geoscience Australia 211
In 2001 the Australian Government launched the Commonwealth Policy on Spatial Data Access
and Pricing, which provided free access to spatial data online, or charge no more than the cost of
transfer for packaged data. This applied to all spatial data collected by government departments
and agencies. Furthermore, in 2009 Geoscience Australia began licensing all data under the
Creative Commons licence, allowing royalty-free use and re-use.
A 2010 report by PwC (‘Economic Assessment of Spatial Data Pricing and Access’) valued
the net welfare gain at $4.7 million. Extrapolating this for all Australian Government spatial data,
Houghton calculated net welfare benefits totally $70 million.
Usage increased rapidly, with the total number of spatial datasets delivered by Australian
government agencies increasing at 112 per cent per annum between 2001-02 and 2005-06, from
around 75,000 to over 1.5 million. Even though the number of datasets rose, intensity of dataset
use increased over the period, with the number of downloads per dataset rising 44 per cent per
annum.
This case reinforces the view that making data available at zero or marginal cost generates welfare
benefits in excess of the revenue lost, with Houghton calculating a benefit/cost ratio of 13.
New Zealand: spatial information in the economy – realising productivity gains 212
An extremely comprehensive 2009 study assessed the value of spatial data to the New Zealand
economy. It concluded that the use and re-use of spatial information added an estimated $1.2
billion to the New Zealand economy in 2008, equivalent to 0.6 per cent of GDP. In addition, it
estimated the non-productivity benefits to be worth many times this figure, through, for example,
planning ‘smarter’ cities and transport systems, aiding national security, or promoting social
cohesion. However, the authors avoided calculating a financial assessment of these benefits
because: there is too much uncertainty around likely events and probabilities and the underlying
value approaches are controversial.
The figures presented by the report should therefore be treated as a ‘lower estimate’ which in
reality could be substantially exceeded.
The report further estimated that barriers to the use and re-use of data cost $481 million of
productivity-related benefits, which would have generated at least $100 million in government
revenue. It recommended releasing the basic spatial information held by the government at
marginal cost. A broader intervention to develop New Zealand’s spatial data infrastructure would
generate a predicted benefit-to-cost ratio of at least 5:1.
In conducting the productivity gains analysis, the report treated historical growth rates of the New
Zealand economy as the base case. The authors then estimated the historical productivity of each
sector without the productivity benefits of spatial information; and the potential productivity if
adoption barriers for spatial information were removed (i.e. the potential unrealised benefit). This
provides an interesting insight into the industry sectors most likely to benefit from the use and reuse of geospatial PSI.
211
212
John Houghton, Costs and Benefits of Data Provision (2011)
Spatial Information in the New Zealand Economy: Realising Productivity Gains (2009)
172
Market Assessment of Public Sector Information
Figure A3.7: impact of adopting PSI
Sector
Quantifiable historical
productivity without spatial
information
Estimated productivity
without adoption
barriers
Crops
1.25
1.88
Bovine cattle, sheep and goats, horses
1.25
1.88
Other animals
1.25
1.88
Raw milk
1.25
1.88
Wool
1.25
1.88
Forestry
5.25
5.71
Fishing
3.44
3.44
Meat products
0.25
0.38
Dairy products
0.25
0.38
Other processed food
0.25
0.38
Coal
0
0
Oil
0
0
Gas
0.63
0.78
Electricity
0.63
0.78
Petroleum & coal products
0
0
Iron & steel
0
0
Other mining
0
0
Nonferrous metals
0
0
Non-metallic minerals
0
0
Chemicals, rubber, plastics
0
0
Wood and paper products; publishing and
printing
0.25
0.38
Textiles and clothing
0.25
0.38
Other manufacturing
0.25
0.38
Water
0.63
0.78
Construction
0.75
1.13
Trade services
0.77
1.15
Transport
2.10
3.15
Communications services
0.82
0.82
Other business services
0.23
0.46
Recreational and other services
0.23
0.46
Government services
0.52
1.04
Dwellings
0
0
Source: ACIL Tasman calculations and estimates
173
Market Assessment of Public Sector Information
Appendix 4: Further statistics
Companies House
Companies House is a statutory body largely funded by fees paid by customers to register a
company. These fees provide the majority of revenue, with users of the data paying for
dissemination costs only, which means the data is free of charge in some cases. Companies
House serves as an ‘information exchange’ – unlike some of the trading funds, it does not provide
value-add services to users, but rather exists to gather data in the form of company registrations
and disseminate this to users.
Figure A4.1 below contains some download statistics provided by Companies House on usage
statistics across its various distribution channels.
Figure A4.1: Companies House usage statistics
Companies House Direct
This is the Companies House online subscription service, for which users pay monthly. It offers free basic
company information as well as company documents and reports for a charge. Usage statistics since 2010/11
are shown below. Companies House report that there were over 405 million free searches between January and
December 2012, indicating high uptake of the free data available.
The number of subscribers to this service rose from 20,967 in 2010/11 to 21,653 in 2012/13 (YTD).
Webcheck
This is the Companies House ‘pay as you go’ service which allows users to check company name availability,
search free basic company information and purchase company documents and reports.
The number of page hits grew from 380 million in 2010/11 to nearly 450 million in 2011/12, with strong growth
posted again for the 2012/13 YTD.
Mobile App
Companies House have released a mobile app which gives free access to information including company
address, status, company appointments and filing history. The number of downloads reported by Companies
House (as of 10/12/12) is 12,362. 11,075 of these were for the iPhone version, and 1,287 for Android. The
iPhone app was released in September 2012 while the Android version was released in November.
URI company searches
Companies House runs a Uniform Resource Identifier service (URI) for all companies on the register. This allows
free searches which will return basic details for the company. Companies House report that since the introduction
of this service in October 2011 it has seen over 111 million data requests.
DVD ROM
Companies House offer a product stored on a DVD ROM, for a fee, containing PDF images of company
documents filed since 31 December 2002. There has been a decline in both one-off requests for the DVDs since
2010/11 reflecting the increased availability of information via the website. One-off requests fell from 931 to 898
between 2010/11 and 2011/12 and subscriptions fell from 664 to 637 over the same period.
Source: Companies House
Land Registry
Land Registry’s main purpose is to register ownership of land in England and Wales and to record
dealings with land once it is registered. The relevant powers and responsibilities are set out in the
Land Registration Act 2002. There are also statutory responsibilities under the Agricultural Credits
Act 1928 and the Land Charges Act 1972.
Under the Land Registration Act 2002, Land Registry has the power to provide ‘consultancy and
advisory services’. The income from these services is used to invest in Land Registry and helps to
ensure that fees for normal registration services are kept to a minimum.
Figure A4.2 below shows usage statistics for the data held by Land Registry.
174
Market Assessment of Public Sector Information
Figure A4.2: Land Registry usage statistics
Land Registry released their Transactional Data in January 2012 and their Price Paid information in March 2012
in open data format. The chart below shows the number of visits and downloads received between their release
and the end of October 2012.
A Land Registry survey of Price Paid Data users between March and November 2012 indicated that the data was
being used to:

Help businesses, for example to provide housing market updates

Assist research, for example into local housing needs and affordable housing policy

Help people buy and sell houses

Help people monitor properties

Present the data in new ways and carry out new analysis
Source: Land Registry
The Met Office
The Met Office uses its data to produce a range of value-added services, some of which are noncompeted services to governments, and others which it sells into competed markets, either to
government or the private sector. Examples of the former include the services provided to defence
and its climate modelling work. Examples of the latter include services provided to aviation clients
and insurance companies. Most paid-for services are therefore value-added, as even wholesale
data requires a high level of processing to enable re-use.
Figure A4.3 contains statistics received from the Met Office on the number of requests and
downloads received by its Data Point service.
175
Market Assessment of Public Sector Information
Data Point
Portal for accessing Met Office reusable data for Web and App developers.
The number of registered users has also increased significantly between December 2011 and January 2013,
from around 100 to over 2000
The Met Office has supplied Deloitte with the number of requests received by data feed in December 2012, which
may be taken as a recent indication of the level of demand – around 2.4 million. The proportions of each request
by data type are shown in the chart below.
As is clear from these figures, the Met Office experiences very high and continual demand for across its data
feeds, with site specific forecasts accounting for over three quarters of the demand, at over 1.9 million requests.
The average daily download of data from the Met Office site, in terms of Mb downloaded, has increased rapidly
over the last six months, as the chart below shows.
The British Atmospheric Data Centre
The British Atmospheric Data Centre (BADC) is NERC’s dedicated data centre for the atmospheric sciences. As
such it has a crucial role in providing long-term curation and access to datasets that support scientific research.
Whilst its main role is to archive the outputs of NERC’s own research activities it also accesses datasets from
many other sources in response to the needs of the research community.

MOHC climate simulations contributing to the CMIP5 archive: 58Tb, multiple models and experiments
176
Market Assessment of Public Sector Information
 MIDAS Observations: 258Gb, all years, all observations
The number of users and the number of files downloaded in 2012 is shown in the charts below.
MyOcean
The main objective of MyOcean is to deliver and operate a single European Ocean Monitoring and Forecasting
system of the GMES Marine Service (OMF/GMS) to users for all marine applications.
OSTIA SST is also made available through the GHRSST portal and has many more users than through my
oceans as registration is not required.
The charts below show the number of unique users and the number of requests in Jan/Feb 2013.
177
Market Assessment of Public Sector Information
Hadley Centre Climate Observation Datasets (HadObs)
Researchers at the Met Office Hadley Centre produce and maintain a range of gridded datasets of
meteorological variables for use in climate monitoring and climate modelling. This site provides access to these
datasets for scientific research and personal usage only. Registered users are a mix of national and international.
It should be noted that one person may be registered for more than one dataset.
Other sources of data include:

Wholesale (ECOMET) Catalogue - The primary objectives of ECOMET are to preserve the free and
unrestricted exchange of meteorological information between the NMS's for their operational functions
within the framework of WMO regulations and to ensure the widest availability of basic meteorological
data and products for commercial applications. There are currently 35 Customers purchasing our
Wholesale data, the majority of whom are resident outside the UK.

WMO Resolution 40 - The meteorological and related data and products described as the minimum set
of data and products which are "essential" to support WMO Programmes and which Members shall
exchange without charge and with no conditions on use.

On-line from the Met Office website – various downloadable data related to content available to view.
Includes historical station data, regional climate data and marine observations.

Weather Observations Website (WOW) – A portal allowing users to upload weather observations, this
also includes the Met Office official monitoring sites. Data can be downloaded based on the content
viewed.
178
Market Assessment of Public Sector Information

Library requests – About 300 requests a year for historical observation datasets of various sizes
provided under personal or research use only terms of use.
It is important to note that the outputs of the Met Office reach many millions of people through channels that are
not represented in these figures. For example, the forecasts of the Public Weather Service broadcast by the BBC
on TV and radio are estimated to reach an audience of between 10 and 20 million per week. The audience
through other channels are estimated at around four million. An additional 5-10 million are estimated to access
forecasts through digital channels such as the web and mobile apps. These, it should be noted, are figures for
‘value-added’ services rather than raw data. The Met Office also provides a large range of other services, many
on a commercial basis, meaning that it is complex to estimate the number of users across the full range of data
and value-added services.
Source: Met Office
Ordnance Survey
In the case of Ordnance Survey, the collection of data is the organisation’s raison d’etre and this
public sector information is therefore not a by-product of other activities. Most revenue, therefore,
is generated from charges levied on users which cover not only dissemination costs but also the
cost of collection. Where value-added services are provided, these are priced to cover the cost of
development. Ordnance Survey products and services are priced at market rates to avoid
undercutting competitors. Access is provided on the same terms for all parties.
The following data was received from Ordnance Survey for this report:
The approximate split between open O/S data and O/S data available under licence and the
number of downloads of O/S data by different channels

For OS OpenData, in 2012 we processed 34,500 orders. Nearly 95 per cent of orders were
for download, the exceptions tending to be for national datasets of OS VectorMap District
and OS Streetview where the volume of data is very high. The top three products by items
ordered were OS VectorMap District, OS Streetview and Code-Point Open. The very
different nature of the products means that a comparison by data volume and number of
downloads does not provide clarity. For example OS VectorMap District may be ordered as
a national set (one order) or as individual 100km x 100km tiles, whereas Code-Point Open
is available only as a national set.
Number of O/S license holders for last three years

We do not keep a periodic record, but at present we have approximately 250 licensed
partners who take the data and on-sell it in value-add products or services. The trend has
been for the number of licensed partners to grow and for the number of direct customers to
shrink, as Ordnance Survey focuses on the key users in specific areas such as Energy and
Infrastructure, while encouraging further development of value-added products by a
growing range of partners
API usage statistics

Below is a graph showing the volume of data on a monthly basis utilised through OS
OpenSpace since January 2011.
179
Market Assessment of Public Sector Information
Source: Ordnance Survey
180
Appendix 5: Empirical methodology
This appendix provides further details of the quantification
methodology used for the current value of public sector information in
the UK. The appendix covers how the report has defined the term ‘value’
and the different modelling assumptions used.
Final estimates and ranges can be found in the main report.
General approach and limitations
As has been discussed in the main report, attempts to accurately quantify the value of current or
future public sector information in the UK are restricted by the lack of data. To recap, all things
being equal, a standard valuation exercise for public sector information would draw upon the
following pieces of data to reach estimates:

an estimate of the total number of public sector information datasets available and not
available;

the relative value of different datasets to one another;

the price paid for different datasets;

how different datasets are used;

the different costs incurred in supplying different datasets;

the number of jobs in the UK attributable to public sector information or involved in its supply or
use/re-use;

the willingness to pay by consumers for different datasets;

the financial value of the different types of benefits (economic and societal) enjoyed by
consumers using and re-using public sector information; and

price elasticities of demand for different datasets and awareness of the shape of consumers’
demand curves.
Further, in order to reach a robust estimate of public sector information value, data or knowledge of
linkages would be required to establish the direction of causality between use and re-use of public
sector information and their impacts in order to properly account for additionality effects.
While elements of the above data requirements are available to differing degrees, the majority of
the above is not available for 2011-12 (the financial year this report focusses on), or for that matter
for earlier years.
In the time available for this research project this report has sought to acquire some of this data
from government officials, PSIHs, consumers and other stakeholders. However, in the absence of
any extensive primary research exercise (including, but not limited to, consumer surveys, data
audits and experiments) there remain significant data gaps. Accordingly, the modelling exercise is
based on a pragmatic approach, making conservative assumptions where necessary and relying
on previously collected data, e.g. on price elasticities. Similarly, the methodology builds on
previous methodologies rather than creating a new one from first principles. The authors of the
Market Assessment of Public Sector Information
report have discussed, and received feedback on, the methodology with the DSB and a number of
other government officials 213 .
The approach has sought to give an indication of the depth and breadth of the value of public
sector information in the UK rather than a definitive statement of its value today or tomorrow. Given
the public sector information is rapidly evolving, we believe it is better to give a broad indication of
the sources and quantum of value (complete with the appropriate caveats) rather than a figure that
has spurious accuracy.
It is acknowledged this approach is not ideal and could be improved with the addition of new
primary research to update assumptions and fill data gaps.
Definition of value
It is important to clearly define what is understand by the term value. From the literature, it is
possible to disaggregate value into four distinct components:

the direct value of public sector information to producers and suppliers (the PSIHs):
these are the benefits accruing to producers and suppliers of public sector information through
the sale of public sector information or related value-added services;

the indirect value of public sector information arising from its production and supply:
the benefits accruing up the supply chain to those organisations interacting with and supplying
PSIHs (but not directly using or re-using public sector information), and the benefits accruing to
those organisations where employees of PSIHs and supply chain organisations spend their
wages;

the direct use value of public sector information to consumers of public sector
information: the benefits accruing to businesses, civil society, individuals and the public sector
from directly using and re-using public sector information for a variety of purposes; and

the wider societal value arising from the use and re-use of public sector information: the
benefits to society of public sector information being exploited, which are not readily captured
elsewhere.
The first three types of value can be grouped as economic value or narrow economic value and
can be measured using standard economic methodologies to derive a monetary estimate for value
and the associated employment figure. The final type of value is harder to measure as it captures
wider benefits to society arising from the use of public sector information – these are typically not
measured in monetary terms. The literature discusses a number of ways in which public sector
information can have broader impacts:

increasing democratic participation: giving citizens and businesses access to public sector
information allows them to perform their own analyses of salient issues, make more informed
choices about public service providers and interact with policymakers to challenge their
assumptions and improve the policymaking process;

promoting greater accountability: for example through the scrutiny of costs of public service
provision and benchmarking comparable services;

greater social cohesion: for example, by providing more information on the provision and
distribution of services, public sector information can be used to dispel myths on who receives
certain public services;

generating environmental benefits: such as reducing congestion and pollution through the
release of better traffic and transport data which helps drivers to better plan journeys; and
213
Although they have not endorsed the methodology used.
182
Market Assessment of Public Sector Information

identifying previously unknown links between different policy areas: through data-mash
ups it may be possible to develop system-wide solutions that holistically seek to address the
root of policy challenges.
Thus, for the purposes of this evidence base, the value of public sector information is defined as
the benefits accruing to the suppliers, users and re-users of the information and data in terms of
profits generated, jobs created and supported and the wider societal benefits.
It is important to clarify that value defined in this way is not the same as the market value of public
sector information. Market value typically refers to the volume of sales multiplied by the price of the
product. In this way it equates to revenues or thus includes labour costs, capital costs, profits and
the intermediate costs of production which are not included in our definition (see below for more
details). Conversely, market value will not capture indirect and induced effects and wider societal
benefits.
Methodologies considered
As highlighted in the literature review appendix, there is no consistent methodology used in
previous research to quantify the value of public sector information which has led to a large
variance between estimates (ranging from £590 million to £16 billion - a factor of almost 30). The
underlying reasons for these different figures appear to be:

whether the quantification approach has been ‘top down’ or ‘bottom-up’;

how the term ‘value’ has been defined;

the underlying data used; and

the category of public sector information being valued.
Below are summarised some of the different methodologies used in the literature.
183
Figure A5.1: Value of public sector information methodologies
Author
OFT / DotEcon (2006)
Houghton (2011)
Koski (2011)
Approach summary
Bottom-up or welfare approach. Value = net surplus
= consumer surplus + producer surplus
Welfare approach and returns to expenditure
approach
Econometric approach examining impact of marginal
cost pricing on real sales growth
Data requirements
 Revenue from PSIHs
 Number of OPSI licences
 Website usage and number of
downloads
 Information on marginal cost
pricing
Assumptions used

Linear demand curves

PSIHs split into three
categories and then subdivided in public sector
information type

Calibrated elasticity estimates

Value-added public sector
information is priced
competitively

For ‘free’ public sector
information, value is estimated
relative to usage of value
added public sector
information

Assume a lower bound of 20%
return on expenditure

Useful life of public sector
information knowledge is 5
years

Elasticity varies over time

Limited to firms within SIC
7420

Average turnover per user
calculated from survey

Ratio of re-users per subdomain was 9.5 (mean) and
8.5 (median)
 Various control variables for firms
Survey of PSIHs asking size of PSI market and sum
of turnover of individual re-users minus cost of
acquisition
 Public sector information turnover
data
Pollock et al (various)
Gains from moving from marginal cost pricing is
2/5(Fλε)
 F = total revenue from sales of
public sector information

Multiplier varies between 5
and 8 and elasticity 2 to 3.5
ACIL Tasman (2009)
Productivity based approach for use and non-use
values – Computable General Equilibrium (CGE)
model
 Extensive macro- and
microeconomic data required for
CGE modelling process

Various assumptions underpin
CGE model
MEPSIR (2006)
 Number of users and re-users
Market Assessment of Public Sector Information
Author
Approach summary
Cebr (2012)
Future efficiency-gain approach – combination of a
literature review tracing impact of R&D and impact
on profits of start-ups due to reduced entry barriers
 ONS economy data by sector
Cowi A/S (2010)
Gains from making address data free of charge
quantified comparing use of data before and after
policy change multiplied by price
 Volumes of data usage
Deloitte Access (2011)
CGE modelling to analyse the productivity improving
from addressing the ‘information glut’
McKinsey (2011)
Combination of an index on value potential and
index on the ease of capture; scaling up drivers in
case studies; and identifying the addressable share
of market for efficiency gains
PIRA (2000)
Investment cost estimates and demand side
estimates built from cost of time, access price, price
paid for public sector information etc.
Data requirements

Assumptions on adoption
rates from literature

Impacts via business
efficiency, innovation and
creation

Counterfactual of what would
have happened anyway
 Extensive macro- and
microeconomic data required for
CGE modelling process

Various assumptions underpin
CGE model
 Spending and taxation data

Various underpinning each
calculation

Two categories of public
sector information: public
sector information for final
users and public sector
information for intermediate
users

Costs allocated accordingly

Assumptions on multipliers
and elasticity

Various
 Graduate and employee data
 Capital stock data
 Bespoke case study data
 Annual reports
 Statistical agencies data
 Proportion of fixed to total costs
PwC (2010)
Welfare approach
Assumptions used
 Administrative costs
 Proportion of purchases by firms
Roger Tym & Partners (2003)
Surveys on willingness to pay and case studies
 Survey of customers
Oxera (1999)
Value-added approach that charts the different ways
public sector information affects the UK economy
 GVA and revenue contribution
185
Following a review of the different methodologies used, two broad ways of quantifying the
current 214 value of public sector information were considered: a bottom-up approach along the
lines of the OFT/DotEcon methodology and a top-down approach using a SIC-SOC matrix.
The top-down approach began by identifying the number of jobs directly involved in the supply, use
and re-use of public sector information. This was on the basis of Standard Occupational Code
(2010) (SOC) categories of jobs and the numbers employed in each. The categories identified as
being involved in public sector information were:

science, engineering and technology associate professionals;

information technology and telecommunications professionals;

IT business analysts, architects and system designers;

IT and telecommunications professionals not elsewhere classified;

management consultants and business analysts;

actuaries, economists and statisticians; and

business and related research professionals.
While this might not capture all jobs involved in public sector information, it is a useful first
approximation. The next stage was to allocate these jobs across different sectors of the economy
using a bespoke SIC 215 -SOC matrix which shows how SOC jobs are distributed across different
sectors of the economy. Having an estimate of the number of public sector information-related jobs
in each sector of the economy allowed the estimation of the level of turnover attributable to public
sector information (using worker productivity ratios) in each sector and then adjust to focus only on
value.
However, as has been noted elsewhere 216 , such a top-down approach, while straightforward, has a
number of drawbacks which reduce its viability. Namely:

it does not properly account for the counterfactual or the additional impact of public sector
information; and

accordingly there is a risk of over-stating the value of public sector information.
While it may be the case that public sector information may underpin a number of products and
services in a given sector, not all the value in that sector can be traced back to public sector
information. Further, such analysis does not explicitly consider what value the sector could
generate using substitutes to public sector information if it were no longer available.
For these reasons, this report uses a bottom-up approach modified from the OFT/DotEcon.
The quantification approach for the current value of public
sector information
The approach has two main stages:

214
Stage 1: estimates of the value of public sector information to PSIHs and value to direct users
of public sector information using a consumer and producer surplus approach; and
The focus was on deriving a methodology for the current value of public sector information rather than the potential
future value in the first instance. As is discussed later, the methodology for quantifying the potential value of public sector
information was more ad hoc given uncertainties over how the market will develop.
215
The Standard Industrial Classification splits the UK economy into over 600 different sectors.
216
For example by the OFT (2006).
Market Assessment of Public Sector Information

Stage 2: estimates of the value of associated indirect and induced impacts to PSIHs using
Input-Output model multipliers (this stage was not carried out by DotEcon).
This report has not sought to quantify the value of the wider societal impacts of public sector
information using a systematic methodology. There is a lack of reliable evidence on the linkages
between the use and re-use of public sector information and the impact on environmental, political,
social and other indices – though, where possible, the report has made ad hoc estimates of the
value generated from particular datasets.
Stage 1: consumer and producer surplus approach – the welfare approach
The welfare approach can be estimated from summing the current net consumer surplus 217 with
the total producer surplus 218 from the supply of public sector information. The approach differs
according to whether the public sector information in question is paid-for or free. By focusing on
consumer surplus, which takes into account the presence of substitutes, the report is able to
accommodate the presence of substitute public sector information datasets and in doing so
captures the additionality impact.
Paid-for public sector information
One of the key drivers for the size of value arising from paid-for public sector information is how
demand varies with price changes (its price elasticity of demand) – this will influence the size of
consumer surplus. When demand is more elastic, the economic value is likely to be lower as
customers will switch to substitute products in response to price rises. When price elasticity of
demand is low or inelastic, economic value will be higher as customers will be less inclined to
switch to substitutes – perhaps because there are few true substitutes.
Clearly, a critical assumption will therefore be the price elasticity of demand for different categories
of public sector information and PSIHs. The value of the price elasticity of demand will depend on
the benefit users and re-users currently enjoy from public sector information in 2011-12 and will
differ by dataset and it unlikely to be stable as new benefits and uses emerge. Additionally, price
elasticity will be influenced by the availability of substitutes and context (e.g. some data is only
valuable at a particular point in time).
In the time and resource available it has not been possible to survey customers to generate new
price elasticities of demand for different public sector information datasets. On this basis, the report
has relied on the long-run elasticities presented in the DotEcon report for the OFT. It is
acknowledged that the elasticity values will have changed since 2006 especially given the greater
availability of open data from PSIHs. However, the other key driver of elasticity – the availability of
substitutes – has not changed significantly since the report as there has been no emergence of
new competitors in the provision of public sector information in most cases. Clearly, this is an area
where new primary research is required, but for the purposes of this evidence base this is
considered to be a sufficient starting point for indicating the value of public sector information.
Following the approach by DotEcon, the report uses three broad classes of elasticity for different
types of PSIHs and public sector information datasets:

Low elasticity: -0.3

Medium elasticity: -0.8

High elasticity: -1.5
217
That is the amount consumers of public sector information are prepared to pay over and above the price they current
pay for access.
218
That is the extent to which revenues exceed the costs of supply – profit.
187
Market Assessment of Public Sector Information
These elasticity figures are treated as the baseline. However, as is good practice, the report
considers adjusted alternative values for each category following research by Davies, Slivinski,
Bedrijvenplatform, Lazo and others 219 .
Figure A5.2: preliminary elasticities assigned to PSIHs supplying pay-for public
sector information
Case
Baseline
Alternative 1
Alternative 2
Low
-0.3
-0.5
-0.2
Medium
-0.8
-1.2
-0.6
High
-1.5
-2.0
-1.2
Source: Deloitte analysis drawing on OFT/DotEcon research
These elasticity figures are then matched to a number of key PSIH suppliers in the dataset.
Figure A5.3: preliminary elasticities assigned to PSIHs supplying pay-for public
sector information
PSIH
Elasticity
HM Land Registry
Low
Registers of Scotland
Low
Companies House
Low
Ordnance Survey
Medium
UK Hydrographic Office
Low
Environment Agency
Medium
Met Office
Medium
DVLA
High
Office of National Statistics
Low
Source: Deloitte analysis
It is acknowledged that these are not the only suppliers of pay-for public sector information in the
UK. However, as noted by Newbery et al (2008), it is estimated that the Trading Funds listed above
comprise around 70 per cent of the estimated total income from UK PSIHs, and this captures a
sufficiently large dataset population (and the subsequent calculations ‘gross-up’ the figure to
capture the full value).
The choice of elasticity assumption assigned to each PSIH is based on a combination of the
understanding of the public sector information held and disseminated by each PSIH, its use and reuse, the level of substitution, discussions with stakeholders and the nature of customers – they
have been adjusted from the 2006 mapping. There is, by necessity, an element of abstraction due
to the diversity of datasets supplied for a fee by each of the above PSIHs.
To accurately calculate total consumer surplus for the pay-for public sector information market, it is
necessary to understand the shape of the demand curves for each dataset. Attempting to derive
individual demand curves for 4,000 plus datasets is clearly not feasible and the report has instead
posited aggregated demand curves for each individual PSIH. For simplicity, and following previous
quantification approaches, the report assumes linear demand curves for pay-for public sector
information. This appears justifiable in this instance along similar lines to those set out by DotEcon:
219
Annexe G Economic value and detriment analysis, December 2006, p.32
188
Market Assessment of Public Sector Information
“While in practice demand curves may not be linear, assuming a linear demand can reasonably be
argued to give a lower bound on consumer surplus, as real-world demand curves are often convex
(as shown in the dotted line) and thus result in a greater difference between willingness to pay and
costs. With more complex specifications of demand, this formula is more complex, but there is still
a one-to-one relationship between revenue and consumer surplus.” 220
There is perhaps more scope for debate as to the shape of the demand curves when it comes to
free public sector information – this is discussed this in more detail below.
Thus, using the assumption of linear demand curves, it is possible to recover the consumer surplus
(i.e. value to the direct users of public sector information) calculated using the standard formula:
Formula (1): Consumer surplus = ½ (Revenue / Price Elasticity)
Figure A5.4: consumer surplus for paid-for public sector information
Source: Deloitte analysis
Formula 1 is used for each PSIH in turn. Revenue data is sourced from annual reports from PSIHs
and adjusted following discussions with stakeholders. In particular, it should be noted that a
proportion of Trading Funds’ revenue is not related to the provision of public sector information and
should accordingly not be included.
Figure A5.5: PSIH revenue and revenue attributable to public sector information
PSIH
Total revenue, 2011-12
(£)
Of which, is public sector information
related (£)
HM Land Registry
360,000,000
72,000,000
Registers of Scotland
58,000,000
12,000,000
Companies House
66,000,000
15,00,000
Ordnance Survey
142,000,000
142,000,000
UK Hydrographic Office
19,000,000
13,000,000
220
Ibid, p.35
189
Market Assessment of Public Sector Information
PSIH
Total revenue, 2011-12
(£)
Of which, is public sector information
related (£)
Environment Agency
417,000,000
834,000
Met Office
196,000,000
121,000,000
DVLA
459,000,000
1,376,000
Office of National Statistics
29,000,000
29,000,000
Source: Deloitte analysis based on Public Data Group publication Approach to Charging and making adjustments for revenue that is not
public sector information related, December 2012. Numbers rounded.
Note the above figures have not been ‘grossed up’ for the missing Trading Funds.
The assessment of producer surplus from paid-for public sector information also follows the
DotEcon approach. As they note:
“In order to estimate the producer surplus or loss, we would require estimates on the costs of
collecting and supplying the data and then compare this to revenue. However, estimates on the
costs of supplying [public sector information] data for each group are not readily available.
In addition, many of the costs of data collection and processing may not be related to the supply of
[public sector information] to the private sector, but might result from other public policy objectives
that already take such costs into account.” 221
Accordingly, when PSIHs do not report a target rate of return on capital employed (ROCE), losses
and profits are assumed to be accounted for in the setting of public policy objectives are left out of
the assessment (i.e. the National Office of Statistics). In contrast, where a ROCE is stated, this
reflects the cost of capital is taken into account when setting public policy objectives and producer
surplus can be estimated as the difference between actual and target ROCE 222 .
Formula (2): Profit = Revenue ([actual ROCE – target ROCE] / [1 + actual ROCE])
Figure A5.6: producer surplus for paid-for public sector information
Source: Deloitte analysis
Again, the report has used financial data from annual reports and stakeholder discussions to
calculate producer surplus (profit) for each PSIH where relevant. For most Trading Funds, the
221
Ibid. p.43.
Following the DotEcon methodology it is assumed all costs are a function of the capital employed and target ROCE.
For further details of the assumptions underlying the producer surplus methodology see Annex G of the OFT report.
222
190
Market Assessment of Public Sector Information
ROCE target is the standard 3.5 per cent figure – actual ROCE data for each Trading Fund can be
found in their annual reports. Given the Office of National Statistics is structured differently and
does not have to make a return it is not included in producer surplus figures.
Free public sector information
Clearly where there is no price for public sector information the above methodology becomes more
challenging – using revenue data to estimate consumer surplus is not feasible as there is no PSIH
revenue attributable to free public sector information. Further there is no producer surplus to
capture as the information is made available for free by PSIHs.
There are also valid questions as to whether the demand curve for free public sector information
datasets can still be assumed to be linear – they may be convex or concave. The different potential
shapes for the free public sector information demand curves are shown below.
Figure A5.7: potential demand curves for free public sector information
Source: Deloitte analysis
In all cases, given the ‘price’ for free public sector information is zero, consumer surplus is the
entire area under the demand curve and can be found taking the integral of the curve.
However, as is known from behavioural economics literature, consumer behaviour can move in
unexpected directions as price moves from zero to a nominal amount and vice versa 223 . In the lefthand corner, assuming a linear demand curve suggests that as price moves away from zero,
consumers react by reducing the quantity demanded proportionally. This may not be the case if
consumers react to the price becoming non-zero by not reducing their consumption proportionately
223
Indeed, one might argue that at zero price the quantity of public sector information demanded should be infinite. This
may not be the case if there are transactional costs involved in the acquisition of the data. For simplicity (and due to the
lack of data) transactional costs are not included in the analysis.
191
Market Assessment of Public Sector Information
– this may suggest a convex, concave or a kinked demand curve. The shape of the demand curve
will therefore have a significant bearing on the size of consumer surplus 224 .
In the available time and resource, it has not been possible to collect primary data on consumer
behaviour to changes in price from zero. In the absence of such data, any specification for the
demand curve will be assumption driven: in the case of linear demand curves that consumers
proportionately adjust quantity demanded to changes in price, in the case of convex demand
curves that the integral sufficiently approximates a discrete summation and is not biased by the
units of scale. Following convention, and as a simplifying assumption (and for reasons of
consistency with pay-for public sector information), the report continues to use a linear demand
curve formulation. However, for completeness an alternative demand curve formulation is
considered.
Beginning with the linear demand curve formulation, in the absence of detailed willingness to pay
data on different types of free public sector information, it is necessary to make a series of
assumptions to recover a provisional willingness to pay for free public sector information.
The DotEcon methodology suggests one approach for this: assuming the choke price (or maximum
willingness to pay) for raw data (taken as a proxy for public sector information) is equal to the
choke price of value-added (pay-for) public sector information minus the current price of valueadded public sector information 225 . Using linear demand curves for modelling simplicity, this
methodology assumes that value-added public sector information is competitively priced in the
main and that the willingness to pay for this would not exceed the willingness to pay for raw data (a
proxy here for free public sector information) plus the cost of value-added features – unless the
consumer would be better of obtaining the raw data and performing the value-add services
themselves.
Using geometry it is possible to recover the choke price for pay-for public sector information.
However, in order to do this it is necessary to use average prices for pay-for public sector
information. This is not straightforward as prices vary by dataset, volume, usage and consumer
type and we have been informed there is no such thing as an ‘average price’ per dataset. The
report has sought to recover a proxy average price using annual report data of Trading Funds and
then used geometry to calculate willingness to pay estimates for different types of free public
sector information. The average value for free public sector information willingness to pay is
estimated to be around £1,300 with the figure varying between £500 and over £2,000. While this
figure may appear high, it should be noted it encompasses the willingness to pay across the
economy and all types of users and re-users, ranging from individuals looking up civil servant pay
levels to SMEs developing apps that harness bus timetables to large multi-nationals downloading
free economic statistics. This appears to be a reasonable first approximation for these indicative
estimates. It would be advisable to conduct more research to ascertain a more accurate,
disaggregated figure.
Having recovered willingness to pay for free public sector information, quantity or usage of this
information is still required. As discussed in an earlier chapter, data on the number of downloads
and page views from a number of major data portals such as data.gov.uk and ons.gov.uk has been
gathered – approximately 2.7 million downloads over a twelve month period. Based on the
assumption that of these downloads around two thirds of the datasets are actually used or re-used
(approximately 1.8 million uses and re-uses). Having quantity used / re-used and the willingness to
pay allows us to generate revenue and using the previous linear demand elasticity assumptions it
is possible to recover consumer surplus for free public sector information.
224
In particular, compare the formula for consumer surplus under a linear demand curve which can be rewritten as
L
c
D
c
CS =1/4p D with the formula for consumer surplus using a Cobb-Douglas convex demand curve: CS =p ln(D) – the
value of consumer surplus in the convex case will be much lower.
225
A fuller description can be found in Annex G of the OFT report.
192
Market Assessment of Public Sector Information
The alternative calculation of surplus assumes a Cobb Douglas demand curve with customers
maximising their utility of consumption from public sector information and all other goods subject to
a budget constraint. Through a series of algebraic manipulations it is possible to recover the
formula for consumer surplus as:
Formula (3): Consumer Surplus = price*ln(demand)
The estimated quantity (demand) and price (willingness to pay) figures can then be used to
calculate consumer surplus in this instance.
The final estimates have taken an average of the linear and Cobb-Douglas estimates.
Stage 2: indirect and induced value of public sector information
Estimates of the value impacts accruing through the business-to-business supply-chain, and
employees spending associated wages, are based upon the UK Domestic Use Matrix for 2005
(latest available tables accounting for import leakage) from ONS and the direct use estimate
derived above.
The direct use estimate of producer surplus (analogous to operating profit) is converted into
expected gross output (GO) for each relevant industry on the basis of information contained in the
UK DUM. This process is uses the inverse of the industry average ratio of operating profit to GO.
GO TYPE I and TYPE II multipliers are then used in an Input-Output setting to consider the
upstream business-to-business purchasing effects (indirect) and consumer spending effects
(induced). The average GO multiplier used in this process was estimated to be in the order of 3.0.
To allow a comparable estimate of value to the original producer surplus (and thus allow
aggregation with other quantified elements), the results are converted back from GO into GVA and
then, in the same manner, operating profit, by definition, providing a comparable surplus estimate.
Again this is achieved by considering the ratios of operating profit to GVA and GO, relative to the
incremental GO for business-to-business spending and consumer spending.
The ‘surplus multiplier’ – the ratio of indirect and induced surplus to the direct use surplus – was
estimated to be 3.4 (1 + 2.4). For each £1 generated as producer surplus in direct use, a further
£2.40 is generated via indirect and induced effects.
Per worker productivity estimates sourced from ONS through the Annual Business Survey are then
used in conjunction with estimates of GVA to provide an indicative level of employment supported
in organisations supplying inputs to the PSI supply chain and supply goods and services to
consumers.
193
Appendix 6: Transport sector case
study
Over the last decade the release of ever greater volumes of detailed and
real-time information has had a considerable impact on the transport
sector. This appendix focuses on quantifying the impact of releasing
transport data in London, where particularly large volumes of
information have been made available. It also discusses impacts further
afield and the future opportunities for increased use of information. It is
therefore an extended version of the transport case studies presented
in Chapter 5.
Overview
Over the past decade the UK transport sector has seen the release of ever greater volumes of
detailed and real-time information. This has allowed travellers to plan and adapt their journeys in
response to real-time updates on conditions across the transport network, as well as allowing
improved public scrutiny of the performance of the transport system.
This section highlights some of the ways in which information has impacted the transport sector.
There is a particular focus on London, which – partly owing to the extent of its transport network –
has seen particularly high volumes of information released, and a correspondingly large number of
products and services developed to allow travellers to access this information and use it to
enhance their journeys. As part of this case study there is an attempt to quantify part of the impact
of releasing this data, by assessing the value of time saved through better access to up to date
travel information.
This case study also assesses the impact of information on the transport sector outside London: in
other cities and on inter-city transport by road and rail. Where possible it offers some estimate of
the value generated by information in these areas, although any such figures should be treated as
high level assessments based on reasonable assumptions, in the absence of the data that would
be needed to conduct more detailed calculations.
Information released by Transport for London
London is unique among the UK’s cities in the extent and complexity of its transport network. Its
underground network alone comprises over 400km of track, with over 500 trains operating across
the network at peak times. Each train travels over 184,000km a year, carrying 1.171 billion
passengers in 2011/12. This is in addition to approximately 19,500 bus stops served by around
7,500 buses 226 ; an extensive overground rail network; light rail networks including the DLR and
London Tramlink; and a riverboat service operating from Greenwich to Millbank. In addition,
London has an extensive road network, with private motor vehicles accounting for more journeys
than the underground and buses combined. 227
226
227
Source: www.tfl.gov.uk/corporate/modesoftransport/1548.aspx
Source: www.tfl.gov.uk/corporate/modesoftransport/2794.aspx
Market Assessment of Public Sector Information
Figure A6.1: TfL rail network map
Source: www.tfl.gov.uk
This complexity means that detailed and current information is vital to enable travellers to efficiently
navigate the network. This includes maps, timetables, and up to date information on closures,
disruptions and delays. Transport for London (TfL) has signed up to the transparency agenda and
provides a wide range of information to users and re-users - principally through its website but also
through other channels. In so doing it builds on the 2010 Mayor’s Transport Strategy, which
included among its commitments “improving the provision of real time and other journey planning
information, including upgrading the TfL web-based journey planner, allowing further improvements
to its real time performance, accuracy and personalisation.” TfL writes that through transparency it
hopes to:

“Enable our stakeholders to hold TfL to account;

Deliver better value for money; and

Enable businesses and non-profit organisations to develop innovative applications using
our data.” 228
Accessing information
The information collected and released by TfL is largely made available to developers and the
general public through their website. For developers, at the time of writing, there are 29 data feeds
228
See www.tfl.gov.uk/transparency/
195
Market Assessment of Public Sector Information
available in TfL’s Developer Area. 229 For information, the complete list of data feeds available is
included at the end of this appendix.
This data is not published under the OGL, and the licensing terms for developers set by TfL include
some restrictions, including branding conditions. The data is, however, available free of charge,
and TfL encourage its re-use in innovative ways, subject to licensing conditions. The TfL website
contains an extensive guidance system providing contextual information and assistance for each
feed, as well as suggestions for its use by developers. 230
For users other than developers, TfL provides information directly through its website under the
‘getting around’ and ‘live travel news’ sections. The information available includes:

a range of maps;

station locations;

accessibility information;

live service updates;

live departure information; and

information on planned works and weekend closures.
Other ways of accessing the information include a free mobile service alerts service for the tube
and DLR. In addition, there are several free tools available for non-developer users, including a
journey planner.
Usage levels
In the absence of usage figures, the analysis here presented is based on the apps available in the
UK through the Apple App Store and Google Play (for Android apps) which use TfL data. This
analysis suggests that in the UK across both platforms there have been around 500,000
downloads of apps relating to bus services, and around 2.6 million downloads of apps related to
tube services. In addition there have been nearly 900,000 downloads of apps covering both the
bus and tube networks. This means that in total there have been nearly four million downloads of
apps using TfL data. 231
It should be emphasised that these figures give only a very rough estimate of the number of users
of TfL data. Many users will access the information directly through the TfL website, through
another website, or through other services. There is also no guarantee that an app download
translates into regular use, as users may download an app but then decide either that they do not
like it (and download an alternative), or fail to integrate it into their daily routine. Nonetheless, these
figures are indicative of the high level of uptake of TfL data.
It is also important to recognise that much of this information has long been available through other
forms. For example, even before the invention of the internet it was possible to access information
about disruptions to travel through the media, such as the television and radio news. Drivers have
long listened to radio traffic updates to help them avoid congestion. However, this information is
now significantly easier and more convenient to access at a time that suits the user rather than the
distributor of the information.
The use of smartphone apps allows instant access to detailed up-to-date information on the status
of all aspects of the transport system, as well as enabling users to plan an alternative route if
necessary. There is likely to be value in this added convenience, both in terms of a greater number
of users of the data, and also due to improved access to the latest information when on the move.
229
See www.tfl.gov.uk/businessandpartners/syndication/16492.aspx
See www.tfl.gov.uk/businessandpartners/syndication/16493.aspx
231
See www.xyo.net; based on a search for ‘London’ within the transit category in the UK site (November 2012)
230
196
Market Assessment of Public Sector Information
Benefits
Conceptually, it is easy to see how the information provided by TfL has generated a range of
benefits. The most notable benefits are likely to be derived from:

Live arrival and departure times for trains, buses and the underground
Developers are now able to offer smartphone apps which allow travellers to see how long
they need to wait for the next bus, or when their next train home will be departing. This
means that they are able to make more efficient use of their time, with less time wasted
waiting, and are able to make informed decisions about the most efficient way of
completing their journey. For example, if there is a 12 minute wait for the next bus, it may
be more time efficient to walk five minutes to the tube station instead. With this information,
travellers are empowered to make better decisions

Traffic cameras and congestion data for the road network
Access to live traffic cameras for drivers and other road users means that it is now possible
to plan a route in advance avoiding areas with heavy traffic and incidents such as road
accidents. This information is available both via the web, as with the BBC Travel News
website shown below, and on the move via smartphone apps. It is also possible to access
information on congestion and average traffic speeds along a route.
Figure A6.2: the BBC Travel News page
Source: www.bbc.co.uk/travelnews/london/trafficcameras

Oyster Card data allowing travellers to track their usage
By registering an Oyster Card online, users are now able to view a history of their journeys
on TfL services, including how much each journey has cost them and the start and end
times. This allows users insight into their usage patterns, including the time and money they
spend on transport, and might potentially encourage behaviour change, for example
changing their travel patterns in order to travel outside peak time and take advantage of
lower fares, where possible.
197
Market Assessment of Public Sector Information

Oyster Card data giving insights into the flows of travellers through the transport
network
A by-product of the Oyster Card system is an extensive database containing details of
every journey taken on the London public transport network, allowing TfL to monitor journey
times and volumes of passengers across the network. This should enable the transport
network to be run in a more resilient way, revealing common bottlenecks and allowing the
impact of delays and closures to be modelled across the network, improving contingency
planning.
That the changes described above will benefit both users and operators of the transport network
seems clear. In many cases it is difficult to identify and isolate the precise impact of each benefit,
beyond incremental improvements to efficiency and user convenience. However, there follows
details of a quantification of some aspects of these benefits, notably time savings to travellers.
Quantification of benefits
TfL ‘lost customer hours’ data
TfL track the performance and reliability of all their transport services. However, each mode of
transport reports its performance in different ways. London Underground publishes performance
reports covering each 28 day period throughout the year. To understand the overall level of delay
caused by a disruption on the network, London Underground uses a measure known as Lost
Customer Hours. This is generated by multiplying the duration of a disruption by the number of
people estimated to be affected, based on the severity of the incident and expected customer
demand at different times and places. Lost Customer Hours therefore accounts both for those who
faced delayed journeys, and for those who were unable to travel at all. Over the year to October
2012 the underground network averaged just under two million lost customer hours per month.
London buses do not provide information on lost customer hours in their performance reports.
Annual lost customer hours have been calculated by taking the figure for the average ‘excess wait’
(beyond that timetabled) and multiplying it by the number of annual passenger trips made by bus.
For private road users, London Streets benchmarks performance by measuring ‘journey time
reliability’ (JTR): the proportion of journeys completed within an allowable excess of five minutes
for a standard 30 minute journey during the morning peak. The analysis is based on taking the JTR
figures and multiplied them by the annual number of driver and passenger journeys, and
separately the number of taxi journeys, to work out the number of journeys experiencing delays in
excess of five minutes. An assumption regarding the average time lost per delay is multiplied by
the number of delayed journeys to arrive at a figure for lost customer hours for private road users.
The approach to rail travel (London Overground) is similar: the reliability figure of 96.6 per cent for
2011-12 from the Travel in London Report and is used to work out the number of delayed journeys.
An assumption regarding the average time lost per delayed journey is then used to calculate a
figure for lost customer hours for London rail travel. Note that this refers only to London
Overground, the TfL-operated rail network. It does not include rail travel to and from locations
outside of London; nor does it include non-TfL services run by national rail operators in London.
Apps using TfL data
An online resource (www.xyo.net) has been used to estimate the number of downloads of apps
based on TfL data, as described above. Although these figures should not be assumed to be fully
accurate, they give a sense of the level of demand for apps using TfL data.
198
Market Assessment of Public Sector Information
Figure A6.3: download figures for apps based on TfL information
Source: Deloitte analysis of data from www.xyo.net
It is unlikely that all app downloads lead to regular use of the app. It may be that a user downloads
an app to try it out, but decides they do not like it and deletes it in favour of an alternative. Equally,
it is likely that some people download an app but then fail to develop the habit of consulting it
regularly. This is accounted for in the model.
The model includes both a conservative scenario, in which 20 per cent of app downloads lead to
regular use, and an optimistic scenario in which 40 per cent of app downloads lead to regular use.
These percentages are then applied to the total app download figures to generate conservative
and optimistic figures for regular app use. Note that the multi-function journey planner apps are
included in the totals for both tube and bus, as they are equally applicable to both. These figures
apply only to bus and tube passengers.
Figure A6.4: scenarios for numbers of users of app based on TfL information
Source: Deloitte analysis of data from www.xyo.net
It should be noted that in addition to the benefit to users of apps (who are second order users of
the data), which is the focus of this analysis, there will be first order benefits from the data through
contributing to an app economy which provides jobs, tax revenue, and other benefits. Even where
apps are not charged for they may create jobs, as there are other mechanisms for revenue
generation (for example through advertising). This ‘infomediary economy’ is an additional benefit
from the release of TfL data. This aspect is not quantified in this analysis, but it should be borne in
mind as another wider economic impact of releasing the data.
199
Market Assessment of Public Sector Information
The model
By making some assumptions about the number of passenger hours saved through better access
to information, and the value of an hour, it is possible to estimate the time potentially saved, and
the value of that time, owing to the information released by TfL.
The model uses Department for Transport figures for the value of time, published in October 2012.
They make a distinction between working time, which is a business cost, and non-working time
(including commuting) which is time to which each individual attaches a value. The value of
working time in the DfT analysis varies by mode of transport. The figures provided by the DfT are
for 2010; in this analysis they have been inflated at the prevailing rate of CPI to estimated 2012
values. The full DfT cost of time figures are provided for reference at the end of this section.
To work out the number of passenger hours saved, the model is based on TfL figures for customer
hours lost, or calculations of these where the data is unavailable, as described above. The
proportion of passengers likely to be using an app based on TfL data to access travel information
is then calculated, based on the download figures.
In addition to the two scenarios regarding the level of regular app use, it is important to recognise
that not all travellers would be able to adapt their route even if they knew that they faced disruption
on their planned route. For example, someone needing to travel from High Barnet to Waterloo
would be likely to have no viable alternative to the Northern Line, even if this meant enduring
significant delays. The model therefore factors in two additional scenarios: a conservative
assumption in which ten per cent of users are able to alter their route in response to information on
delays, and an optimistic scenario in which 25 per cent of users are able to alter their route.
Although DfT provide a range of values of time, this model uses the lowest value – for
leisure/commuting – rather than the higher working time values. Adopting this approach generates
conservative estimates, in the absence of more detailed information on the types of travellers
affected by delays. It is also reasonable to assume that the majority of journeys do not occur with
business as the primary journey purpose. However, it should be noted that introducing the value of
working time into the model would mean that the estimated value of time saved would become
significantly higher due to the large differential between leisure/commuting and working time in the
DfT values of time.
Findings
The estimate of the value of time saved through use of TfL data generated by the model ranges
from around £15 million to nearly £58 million per annum (based on 2012), depending on the
scenario for level of app use and ability of travellers to change their journey routes.
Of this value, around a third is accounted for by savings for underground passengers, and another
third by savings for passengers of London trains (over-ground). The remainder is split roughly
equally between bus passengers and private vehicles. Taxis account for a relatively small level of
savings, due to the low number of passengers relative to other modes of transport. It should be
noted, however, that if the values of working time are used the savings for taxi users become
relatively more significant, as the DfT accords taxi users the highest value of time of any mode of
transport.
These figures are shown in the chart below.
200
Market Assessment of Public Sector Information
Figure A6.5: the value of time saved through use of TfL information, 2011/12 (£
million)
Source: Deloitte analysis
The number of beneficiaries, i.e. passenger journeys which save time as a result of travellers using
TfL information, is shown in the graph below (millions of journeys). Under the most conservative
scenario around 200 million passenger journeys per year are estimated to save time due to TfL
information, while under the most optimistic scenario the figure is over 700 million passenger
journeys. The highest number of beneficiaries are private car users, indicating that under this
modelling approach there are a large number of car users experiencing savings to their journey
time, but that the time saving to each journey is relatively small.
201
Market Assessment of Public Sector Information
Figure A6.6: the number of passenger journeys experiencing time savings through
use of TfL information, 2011/12 (millions of journeys)
Source: Deloitte analysis
Summary
It is important to note the limitations of these figures. They attempt to model only time saved due to
travel disruption avoided – they omit the time potentially saved in everyday travel, for example by
allowing commuters to time their exit from the office so as to catch the next bus.
They are furthermore based on a range of assumptions regarding traveller behaviour, specifically
assumptions regarding app use and ability to alter their route of travel, as well as estimations of the
value of time that may not be accurate for all travellers.
For these reasons the approach here adopted is the most conservative that seems reasonable.
The figures represent the lower bound impact of TfL data – in other words, the true impact is likely
to be significantly higher. By way of comparison, the much fuller HS2 benefit assessment
estimated annual time saving benefits at a net present value to 2043 of £7.3 billion. 232 On a per
annum basis, over the 17 years between 2026 and 2043 (before which the line is not scheduled to
be operational) this equates to £417 million in annual journey time savings, or £440 million in 2012
prices. 233
232
See http://highspeedrail.dft.gov.uk/sites/highspeedrail.dft.gov.uk/files/hs2-economic-case.pdf
There are a number of shortcomings to this estimate, notably that it calculates the benefits on a straight line basis.
This report has not conducted an in depth analysis of the benefits of HS2: the figure is included in order to illustrate a
point, which is that even making conservative modelling assumptions, the annual time saving benefits delivered by
233
202
Market Assessment of Public Sector Information
This is based on many fewer passengers but a much greater journey time saving per journey. As
such, the provision of exhaust PSI in London and its environs, at what is a relatively low marginal
cost to Government, may create as much as 13 per cent of the annual monetised time savings of a
major infrastructure investment such as HS2. In short, information is time, and time is money.
Finally, these figures do not take account of the broader welfare benefits – for example social and
environmental – which is likely to results from reducing friction and improving efficiency for
travellers around London. These benefits, although beyond the scope of this study and difficult
quantify, are likely to be significant.
TfL statement
In the course of the work on this case study, discussions were held with TfL. They have provided the following
statement for inclusion in the report.
“Every day, millions of Londoners and visitors rely on the information provided by Transport for London (TfL).
Our online journey planning and service update tools are essential sources of travel advice used by 8 million
customers. And through traditional media such as radio, television and newspapers, our travel information
reaches millions more.
But in our aim to deliver world leading customer service, we are now realising the enormous potential of
opening up our data sources to the wider community. Our digital strategy, formulated in line with the UK Open
Government agenda and the Mayor of London’s open data policies, sets out our commitment to free data,
updated in real time where possible, to encourage web and app developers to create the tools and services our
customers want. Over 5,000 developers have already signed up to our data feeds, supporting hundreds of
travel apps and helping millions of end users – all achieved at much less cost than were the same services
created using TfL’s own resources, and while supporting the digital economy.
The wider economic benefits of releasing data are clear. If transport is disrupted for some reason, real-time
customer information can alleviate the impact by helping people choose alternative routes. The case study
presented here estimates that this benefit alone may be worth tens of millions of pounds per year. Yet even
when transport is running smoothly, TfL’s journey planning information helps people select the quickest or least
congested route, while recent innovations, such as our release of real time bus arrival data, enable customers
to adopt a ‘just-in-time’ approach to their travel, reducing waiting time and freeing up time for more productive
activities.
TfL has also exploited the opportunities of customer information, both open source and through our own
channels, to tackle some of the biggest challenges a transport operator can face – from alerting people to
travel hotspots during the London 2012 Games, when 60,000 people followed dedicated Games travel advice
on Twitter, on top of TfL’s usual 400,000 followers, to supplying real time service updates to keep London
moving during critical transport infrastructure upgrades. Moreover, it’s not just TfL’s customers that benefit.
Providing high quality real time information also helps us, as a transport operator, to provide a better service
through people avoiding already crowded stations or roads, and to recover quickly from disruption by giving our
customers choices.
While providing substantial benefits, the continued release of data does incur some up front and ongoing
management costs, so we would encourage developers using our feeds to share analytics with us on how
customers use their products, to improve our ability to release the most useful data and drive the greatest
value for money.
In summary, we recognise the substantial advantages resulting from open transport data, for customers, for
operators, and for the wider economy. We welcome the findings of this case study into the impact for disruption
mitigation, and will continue to seek ways of making our data even more accessible and useful.”
releasing transport data for London are a significant proportion of the estimates for journey time savings delivered by a
major national infrastructure project such as HS2, for much lower investment of resources.
The estimated savings for HS2 include values of working time. If only the value of leisure/commuting time is included (as
is the case for the calculations of the value of savings in London), the value of time saved by HS2 falls to £105 million per
year. This figure is not used as it is likely that HS2 will be used more heavily for business travel, whereas London
transport is more heavily used for commuting and leisure purposes. However, this illustrates that the relative significance
of the savings due to release of information are likely to be greater than the 13 per cent of HS2 savings suggested
above.
203
Market Assessment of Public Sector Information
Information relating to transport beyond London
This case study focusses on London and specifically the information released by TfL. London
provides a particularly compelling example of the benefits that can be generated by releasing
information. This is partly due to the large volumes and variety of information released by TfL, and
the user-friendly formats in which it is provided. It is also because London is unique in the UK both
in its size as a conurbation and the scale and complexity of its transport network.
It is reasonable, however, to anticipate that benefits from the release of information in other cities
in the UK also exist, either actually or potentially; as well as benefits to inter-city rail and road
transport. Anecdotally, the transport systems of other major UK cities such as Manchester and
Leeds are not well served by tools such as smartphone apps, either because the information has
not been made available or is not in an adequate format, or because there are not a sufficiently
high number of users to justify creation of an app. The latter seems unlikely, given the number of
apps available and the very limited use base to which some of them cater: if nothing else, one
would expect an app to be created as a service to the community.
An example of an initiative in this area is the Future City Demonstrator, for which Glasgow won £24
million of government funding early in 2013. In its press release the ODI noted that “open data will
be at the heart of the programme. People will be able to monitor traffic levels on the road, before
beginning their journeys. They will also be able to check whether bus and train services are
running to time.” 234 In many ways this is similar to the services already available in London,
indicating that other cities are interesting in developing a similar ‘open data infrastructure’ to
support and improve the operations of their physical transport infrastructure.
Due to smaller transport networks and fewer passenger journeys, the scale of the benefits realised
are unlikely to match that experienced in London. For example, the Glasgow subway contains 15
stations and a route length of ten kilometres, just 2.5 per cent of the length of the London
Underground. Nonetheless, all things being equal it is to be expected that there will be benefits in
line with the scale of the transport systems involved. There will also be a broader social and
environmental value to the more efficient running of these transport systems.
In addition to benefits in other cities, there are potential benefits to be realised in inter-city
transport. With nearly 490 billion vehicle kilometres recorded on major roads in England and Wales
in 2011, and just 81.9 per cent of journeys completed on time in the year to October 2012, there
appears to be significant scope for improvements. 235 The case study below highlights one example
of how the release of information is capable of improving inter-city transport.
Traffic England
Traffic data is available through Traffic England, a service provided by the Highways Agency’s National Traffic
Information Service. The information covers most of the motorways and major A-roads in England. Data
includes real time information on traffic speed for each section or motorway or road, details of disruptions and
closures, and other useful information such as weather conditions. Users are able to employ the information
provided through this service to:
 monitor regular commutes;
 avoid unnecessary queues and delays;
 see how busy the roads are by viewing live traffic cameras;
 identify roadworks and whether or not they are causing delays; and
234
See www.theodi.org/news/glasgow-wins-%C2%A324m-boost-after-recognising-%E2%80%9Crevolutionaryimpact%E2%80%9D-open-data
235
DfT road congestion statistics
204
Market Assessment of Public Sector Information
Traffic England
 plan business trips, deliveries and holidays by checking future roadworks, events and forecast traffic
conditions that are expected to cause delays.
Traffic England - Motorway Traffic Flow
Source: www.trafficengland.com/motorwayflow.aspx?ct=true#mtf
Traffic England notes that the website receives around 960,000 visitors per month. 236 As an indicative
estimate, if 50 per cent of these monthly users were to save ten minutes of journey time, using a low value of
time of £6.80 per hour (based on DfT figures), this would generate time savings worth over £544,000 per
month, or over £6.5 million per year. These are very high level estimates based on broad assumptions about
the level of time savings achieved by users, but provide an indication of the scale of benefits, in terms of time
saved, that could be expected to be currently generated by this service.
It seems safe to assume that not all road users who could benefit from this information are currently aware of
or using it. It is therefore likely that the potential or ‘latent’ benefit from increased exploitation of the information
is even greater than the current benefit in journey time savings. In 2011 there were nearly 490 billion vehicle
km travelled on major roads. 237 If even one per cent of these could made more efficient or timely through the
use of better information, this would improve journeys totally nearly five billion km. Assuming an average speed
of 35 km/h, this would equate to 140 million hours of journey time affected. Note that this is not time saved, but
serves to illustrate the potential scale of the impact of improved access to information to road users if even very
conservative assumptions are applied.
In addition to the financial value of time saved, there will be other financial savings, including fuel savings, and
avoidance of cost to delivery services and the businesses they serve owing to delays. There will also be wider
non-financial benefits, including environmental benefits from reduced transport emissions, and benefits arising
from reduced road congestion.
Further to the ‘latent’ benefits that could be gained from increased use of the available information under
current conditions, there is likely to be an additional benefit from future use as congestion becomes more
prevalent across the UK road network. The DfT estimates that in 2010 eight per cent of traffic travelled in ‘very
congested conditions’. It forecast that by 2035 this figure would be 17 per cent, rising to 42 per cent for London
(if no further investments in road infrastructure are made beyond those set out in the 2010 Spending review).
This equates to 32 lost seconds a mile for all traffic by 2035, and 140 lost seconds a mile in London. 238 This
increased risk of congestion and delays seems likely to increase the need of drivers for detailed and up to date
236
See www.trafficengland.com/faq.aspx#a28
See www.gov.uk/government/publications/road-traffic-estimates-quarter-2-2012
238
See http://assets.dft.gov.uk/publications/road-transport-forecasts-2011/road-transport-forecasts-2011-results.pdf
237
205
Market Assessment of Public Sector Information
Traffic England
traffic flow information, and increase the amount of travel time that could potentially be saved through use of
information.
In addition to increased exploitation of the information, it is likely that its value to road users will increase as
more information is made available, and the tools for accessing and manipulating the information become more
powerful. Traffic England is currently running a Traffic Map Beta which offers improved and more resilient
information on speeds by drawing data from a wider range of sources, as well as extended coverage. As the
information improves it is likely to generate higher levels of value for users.
One of the challenges for the release and use of information relating to the national road network is
the lack of cohesion amongst the responsible bodies. The House of Commons Select Committee
on Transport has summarised the organisational structure:

the Secretary of State has responsibility for overall Government policy on roads, puts the
relevant legislation in place, sets the strategic framework for new developments in traffic
management, and establishes financial parameters;

the Highways Agency is an executive agency of the Department for Transport (DfT) and, on
behalf of the Secretary of State, operates, maintains and improves the strategic road
network - most motorways and all-purpose trunk roads - in England;

local highway and traffic authorities - County Councils, Metropolitan Borough Councils,
Unitary Authorities, London Boroughs and Transport for London - are responsible for all
other public roads (including non-trunk 'A' roads, 'B' and 'C' roads) and a small number of
short, motorway standard 'A' roads in major urban areas; and

Integrated Transport Authorities (ITAs) (which replaced the six English Passenger
Transport Authorities in 2009) have full responsibility for local transport plans in their cities
and can modify governance arrangements within their areas. 239
Given this complexity, information is often split across different jurisdictions and may be
unavailable in a consolidated form or a single location – unhelpful for road users whose journeys
may well take them across several administrative boundaries. The case study below examines the
role of roadworks.org in improving access to data by combining data held by multiple authorities.
Improving access to fragmented information – roadworks.org
Information on when and where roadworks will take place has the potential to allow road users to plan their
journeys better, in order to avoid congestion and the inconvenience and economic cost this incurs. However, in
England and Wales roadworks information is split across 175 local Highways Authorities. Where it is published
by local authorities this is done using a variety of bespoke platforms, which are costly to administer and cover
only the area within the administrative boundaries of that local authority. The diverse formats of this
disaggregated information make it difficult to incorporate into real-time systems with national coverage. This
situation therefore incurs higher than necessary costs for local authorities while failing to serve the needs of
road users.
Roadworks.org is an attempt to provide a solution to this problem. It currently publishes details of over two
million roadworks annually, from over 140 Highway Authorities in England and Wales.
239
See www.publications.parliament.uk/pa/cm201012/cmselect/cmtran/872/87204.htm#a1
206
Market Assessment of Public Sector Information
Improving access to fragmented information – roadworks.org
Roadworks.org
Source: www.roadworks.org
This aggregated national database has demonstrated significant benefits including:

Many utility companies and their contractors, including BT Openreach, have embedded roadworks.org
within their works management systems, enabling pre-coordination of planned works with those local
authorities that have implemented roadworks.org. This reduces clashes and the disruption and
congestion caused by works.

Roadworks.org provides a data feed API to over 60 companies, enabling them to use up to data and
broad coverage roadworks data to develop innovative services.
A recent report by ELGIN estimated the total benefits at £25 million per annum. This includes tangible benefits
of £6.3 million arising from costs savings to each local authority owing to the greater efficiency of the service as
compared to individual bespoke systems.
Tangible savings
Savings (£ per LA)
Better coordination
15k
Communication with stakeholders
20k
Public enquiries
7.5k
Duplication of systems
25k
Fixed Penalty Notice revenue
5k
The remaining £19 million savings are calculated as ‘intangible savings’ from the benefits of reduced
congestion owing to better coordination of works; and greater operational efficiency for utilities companies and
local authority Highway Maintenance departments. These estimates are likely to be less accurate but give an
indication of the scale of the benefits achievable.
There are costs associated with this system. The paper calculated the total costs of operating the
roadworks.org system to be £700,000 per year (based on 2011/12 costs), implying a subscription fee of £4,000
per local authority if all 175 local authorities were to subscribe. This is set against estimated tangible savings of
£72,500 per local authority, in addition to the intangible benefits, meaning a net tangible saving of £68,500 per
local Highway Authority (or £12 million p.a.).
207
Market Assessment of Public Sector Information
Improving access to fragmented information – roadworks.org
This case demonstrates an important point regarding information: there is generally a cost to making it
available in the most useful form for users and through an effective distribution channel. However, where the
cost benefit analysis is favourable, this is an investment worth making – even before the additional downstream
benefits are estimated. This case also demonstrates the role that the private sector can play in helping public
sector organisations make the best possible use of their data, in order to derive the maximum benefit for the
lowest cost. It is also a unique example of governance to protect the a critical national dataset which is Open
but crated by private risk capital
Source: Elgin, ‘A new public-private model for creating a national database of local roadworks’ (March 2013)
There are also benefits to be gained from opening up data in the rail industry. There is
considerable effort currently being expended to increase the availability of data in this space, as
the case study below demonstrates.
Disruptive innovation in rail travel – Placr and other providers
Several companies are focussing on using information to disrupt the rail industry through innovative products
and services. The objective is a liberalised data market which will allow a range of value-added services to
arise, with the potential to considerably enhance the customer experience and disrupt the industry status quo –
either through providing more accurate information on the nature of the service and any delays, cancellations
or route changes, or by allowing customers to find more competitively priced ticket options.
It should be noted that much of the data discussed in this section may fall outside the strict definition of public
sector information. The train operating companies are private, while Network Rail, as a statutory corporation,
operates the physical rail infrastructure. This means that the status of data relating to privately operated trains
running on a quasi-public network data is unclear. Given, however, that the rail industry receives billions of
pounds of government subsidy annually, it could be argued that there is a public interest in all data relating to
the rail industry being treated as public sector information. In theory, therefore, the government could make
provision of information by the train operating companies as if it were public sector information a condition of
operating a franchise on the network. For the purposes of this analysis we treat all information discussed as
public sector information. We consider that to omit this information would be to ignore a vital part of the picture.
One company operating in this space is Placr, a start-up focussed on location and transport data, currently
being incubated by the Open Data Institute. Its website states: “Our chief objective since foundation has been
the creation of a single UK source of transport information by unification of timetable, live departure and
disruption information for bus, rail, metro and ferry services.” 240
Barriers
Placr has identified a range of both opportunities and barriers. Industry data streams that they would like
access to include:

short term cancellations;

rolling stock formation, i.e. how many carriages does a train contain; and

cycle policies.
In addition, data that is made available by the Association of Trade Operating Companies (ATOC) under the
Rail Settlement Plan incurs a charge for use, including a license fee of £5,005 annually for data supplied daily
or weekly, and an annual quoted charge of up to £27,430 for daily fares, timetable and routeing guide data. 241
This is sufficiently high to act as a potential disincentive for a start-up that has yet to generate a reliable
revenue stream. ATOC does provide trial data for proof of concept and test purposes, but this is not
necessarily up to date and therefore may be inappropriate to be used for commercial purposes.
Licensing conditions for rail data are also said by some stakeholders to be confusing for developers and have
been used to shut down services. For example, in 2009, in following a dispute between Kizoom and ATOC,
the latter withdrew Kizoom’s licence to use to use train department information for its free MyRailLite app 242
240
See http://placr.co.uk/history.php
See www.atoc.org/about-atoc/rail-settlement-plan/data-feeds
242
See https://mocko.org.uk/b/2010/10/29/national-rail-have-killed-my-train-times-app/
241
208
Market Assessment of Public Sector Information
Disruptive innovation in rail travel – Placr and other providers
Developers may therefore face some challenges in accessing data in the rail industry.
In terms of the opportunity, Placr identify three areas in which they believe increased availability of data can
disrupt the rail industry and improve the experience of passengers.
Providing live service updates
Increased availability of data has the potential to allow passengers current information on the status of rail
services. Some of this information is already available through the National Rail website, TOC websites and
mobile apps, which provides live departure boards and allows users to track a train’s progress throughout its
journey.
National Rail Enquiries
Source: http://nationalrail.co.uk/times_fares/ldb/
There are other, potentially more interesting ways of tracking service quality. For example, the train lateness
map available at transportapi.com represents ‘lateness’ as coloured bubbles on a map, allowing users to
intuitively see where the delays are and plan their journey using the best-performing route.
243
See www.whatdotheyknow.com/request/atoc_written_advice_to_dft_on_im
244
For example, see http://www.dailymail.co.uk/news/article-2271415/Earl-Attlee-defends-train-fares-admitting-expensive-forcedcancel-day-motor-show.html
209
Market Assessment of Public Sector Information
Disruptive innovation in rail travel – Placr and other providers
Train lateness map
Source: www.scribd.com/doc/123365071/Friday-Lunchtime-Lectures-at-the-ODI-How-can-Open-Data-Revolutionise-your-RailTravel
By presenting the data in innovative formats, service users are able to interact with it in new ways and gain
additional insight into the performance of travel services and the best way to complete their journey. Travel
becomes more intuitive and users gain enhanced understanding of the travel options at their disposal, which
should translate into faster and more efficient journeys.
Generating financial savings for passengers
It is hoped that fare data will soon be made open. This data, and the innovative services that arise from it, is
likely to be of benefit to passengers. Part of the potential value of this data to passengers has been
demonstrated by the example of TfL, which now allows holders of a registered Oyster Card to view their
journeys online, and also offers to send monthly statements detailing all journeys and credit top-ups during the
period. This gives users the ability to track their expenditure patterns on public transport, and also allows them
to identify any journeys where they believe they have been wrongly charged and apply for a refund. The data is
therefore of clear benefit to users.
New and potentially disruptive services may also be built on this data. An example of using data to generate
savings for travellers is Tickety Split, a service run by Money Saving Expert. This service is based on the
insight that buying separate tickets for constituent parts of a journey is often cheaper than buying one ticket.
Although the journey may take longer and require changing trains, it may be that the traveller values the
financial saving more highly than the opportunity cost of additional time spent on the journey. Train operating
companies do not generally offer their customers this choice, and searching manually for the best split ticket
combinations can be cumbersome. Tickety Split searches available fare data to calculate the cheapest
combination. In the example below, a ticket from London to Leeds costs £97, but this can be reduced by
£13.30 to £83.70 if intermediate tickets are bought from London to Grantham and from Grantham to Leeds.
210
Market Assessment of Public Sector Information
Disruptive innovation in rail travel – Placr and other providers
Tickety Split
Source: http://splitticket.moneysavingexpert.com/
Information on the impact of fare-splitting on revenues is not publically available. 243
It is, however,
possible to roughly estimate the scale of the savings that might be achievable through this technique, based on
the following:

franchised passenger revenue in 2011/12 was £7.229 billion;

around 75 per cent of this revenue was generated by ordinary fares (assuming that season ticket
holders are unable to secure savings through ticket splitting); and

assuming an average saving through ticket splitting of 13 per cent.
Based on these assumptions, passengers could save an estimated £705 million per year through ticket
splitting. The full extent of these estimated potential savings are unlikely to be realised, due to varying levels of
awareness and adoption among passengers. Even if only 25 per cent of passengers were to take advantage of
split ticketing, this would still equate to an estimated £176 million saving per year. This indicates that
increasing accessibility of fare data holds large potential benefits both for individual passengers and for the
travelling public overall. In addition to the financial gains to passengers, this example also demonstrates the
ways on which data can be used to empower consumers. In so doing it is likely to induce train operating
companies to run more efficient, consumer-focussed operations and to offer more competitive fare pricing.
Future releases of data could offer the potential for additional innovative services which would increase
potential savings to passengers. For example, the data from a smart ticketing system such as Oyster could be
linked to real time data on train movements to provide a service whereby automatic refund claims are filed on
behalf of passengers when their trains are delayed. This could have a genuinely disruptive effect on the
industry, forcing train operating companies to focus on delivering a reliable service, as a poor level of service
would have a direct impact on revenues. Were such a service to be developed, it would depend on access to
data that can be used to monitor reliability.
Creating independent streams of data about rail network performance
Many of the benefits to users discussed above will depend on data that can provide an independent view of the
performance of the railways. For example, Tube Radar (based on Open Street Maps) provides a view of
performance (time intervals between trains) across the network, as compared to normal performance. Placr
also cite the example of being asked by the press to offer a view on levels of service during a recent strike, in
211
Market Assessment of Public Sector Information
Disruptive innovation in rail travel – Placr and other providers
the face of conflicting reports from TfL and the unions.
Tube Radar
Source: http://tube-radar.com/
One source of this data is smartphones carried by passengers, which can be used to monitor the progress,
frequency, acceleration/deceleration and other details regarding train operations. By crowd sourcing this data,
an additional window on the performance of the rail network is opened up, increasing pressure on train
operating companies to improve standards and provide a high quality service to their customers.
Conclusion
Placr and other start-ups operating in this space offer numerous ideas for products and services that could
collectively transform the experience of rail travel in the UK, by empowering passengers and enabling them to
hold rail operating companies to account. As average rail fares rise year on year, with many users perceiving
rail travel as offering poor value for money,244 data offers a mechanism ensuring that service users and
taxpayers receive value for money from train operators.
Some of the data required to drive this transformation remains either unavailable or subject to restrictive pricing
and licensing conditions, although there are signs that this situation may gradually be changing. There is
demonstrable value to be generated from releasing this data, both in terms of financial savings to users,
convenience in journey planning, and in driving accountability (and from thence efficiency) in the rail industry.
While not strictly public sector information, given the level of taxpayer contribution to the rail industry and its
status as infrastructure of national importance, there is a clear responsibility for government to ensure that the
industry operates in an open and accountable manner. This seems likely to be an area in which increasing the
openness of data will generate rapid and measurable returns to rail service users, taxpayers and the wider
economy.
Source: much of the content of this section is derived from a session at the ODI, “How can open data revolutionise your rail travel?”
(01/02/2013). Slides available at http://www.scribd.com/doc/123365071/Friday-Lunchtime-Lectures-at-the-ODI-How-can-OpenData-Revolutionise-your-Rail-Travel
Figure A6.7: additional data 1
Syndicated feeds available to developers through the TfL website
212
Market Assessment of Public Sector Information
Syndicated feeds available to developers through the TfL website
The full list of syndicated feeds offered by TfL is included below. TfL state “before we give permission to use
any feeds, we need to know how they will be used, where they will be used and how many people are likely
to view them.” Gaining access to these feeds requires:

Providing personal and/or company contact details

Providing information on intended use, target audience and estimated audience numbers

Agreeing to the terms and conditions
The full list of feeds available is provided below (correct as of January 2013).

Live traffic camera images v2

Live bus arrivals API (instant)

Source London Charge Point data dictionary

Journey Planner API Beta

Tube station accessibility data

Rolling origin and destination survey

River services timetable

Tube departure boards, line and station status

Journey Planner Timetables

Coach Parking sites/timetables

Licensed private hire operators – Find-a-ride

Pier locations

Bus routes

Tube – this weekend

Station facilities

Live bus arrivals API (stream)

Live Roadside Message Signs v2

Source London Charge Point Location Data

Barclays Cycle Hire availability

London Underground passenger counts

Public transport accessibility levels

Barclays Cycle Hire statistics

Oyster card journey information

Live Traffic Disruptions

Dial a Ride statistics

Station locations

Oyster Ticket Stop locations

Bus stop locations

Tube – this weekend v2
Source: www.tfl.gov.uk/businessandpartners/syndication/16492.aspx
Figure 6.8: additional data 2
213
Market Assessment of Public Sector Information
DfT value of working time per person (£ per hour)
Vehicle occupant
Market price (2010
prices and values)
Market price (2012 prices
and values – Deloitte
analysis)
Car driver
33.74
35.4
Car passenger
24.17
25.4
LGV (driver or passenger)
13.00
13.7
OGV (driver or passenger)
13.00
13.7
PSV driver
13.00
13.7
PSV passenger
25.81
27.1
Taxi driver
12.47
13.1
Taxi/minicab passenger
57.06
59.9
Rail passenger
47.18
49.6
Underground passenger
45.90
48.2
Walker
37.83
39.7
Cyclist
21.70
22.8
Motorcyclist
30.53
32.1
Average of all working persons
34.12
35.8
Figure A6.9: additional data 3
Appendix 3 – DfT value of non-working time per person (£ per hour)
Purpose
Market price (2010
prices and values)
Market price (2012 prices
and values – Deloitte
analysis)
Commuting
6.46
6.80
Other
5.71
6.00
Source: Department for Transport, ‘TAG Unit 3.5.6: Values of Time and Vehicle Operating Costs’, October 2012
214
Appendix 7: Further case studies
This appendix contains further details on the healthcare and public
sector case studies outlined in Chapter 5.
The healthcare and life science sector
This section considers examples of how the release of public sector information datasets relating
to the healthcare sector can have a beneficial impact. This may be in terms of identifying new
efficiencies in the NHS, or it may be in improving patient outcomes, which may itself result in cost
savings either directly to the health service or to the wider economy.
Mastodon C – identifying NHS prescription savings from big data
Case study summary
By using data on prescribing practice across England, variations in spending on different classes of drugs can
be identified. It is then possible to calculate the potential savings to be achieved by moving from prescribing
branded to generic drugs.
Size of the prize

For statins alone, the NHS could save around £200 million per year by reducing prescriptions of branded
in favour of generic versions.
 When extended to all classes of drugs, the total potential savings could amount to £1.4 billion per year.
NHS prescribing data is released at GP level through the Health and Social Care Information
Centre. These are large datasets, with around 10 million lines of data released every month – an
example of ‘big data.’ Mastodon C, a start-up company currently being incubated by the Open Data
Institute and describing itself as an ‘agile big data specialist’, saw an opportunity in this data to
identify potential efficiency savings for the NHS.245
Prior to this analysis there was already an awareness in the sector of the possibility of achieving
savings through changes to prescribing practices. The British Medical Journal has published
research indicating that the potential savings to the NHS of switching from branded to cheaper (but
in many cases equally effective) generic drugs could total £1.4 billion246. There have been previous
and ongoing attempts to achieve savings through increasing prescription of generic drugs, notably
through the work of prescribing advisors. However, these attempts have hitherto met with limited
success. GPs may habitually prescribe a branded drug without considering the cost implications;
equally, patients may be accustomed to a branded version and feel that a generic is an inferior
substitute. For these and other reasons, it has proved difficult to bring about significant change in
behaviour.
Mastodon C, working with Open Healthcare UK, wanted to try a new approach to raising
awareness. They used big data to demonstrate regional differences in the cost of prescriptions,
hoping thereby to drive change by allowing GPs and PCTs to compare their performance to those
of GP practices and PCTs across England. By highlighting seemingly unwarranted variation, GPs
and PCTs could identify where they were underperforming compared to their peers in terms of
keeping prescribing costs as low as possible.
245
Source: Mastodon C presentation at the ODI, 18th January 2013. Slides available at
www.scribd.com/doc/122981341/Friday-Lunchtime-Lectures-at-The-ODI-Big-Data-Comes-to-the-NHS. See also
http://blog.mastodonc.com/ and http://prescribinganalytics.com/
246
See www.bmj.com/content/341/bmj.c6449
Market Assessment of Public Sector Information
The team chose statins as the focus of their project, because there is widespread agreement that
the generic version is in most instances as effective as the branded versions (based on the
guidelines provided by NICE). The price difference is also significant: £1.30 for a generic version
as opposed to £20 or more for many branded statins, as shown in the Figure A7.1.
Figure A7.1: prices of generic and proprietary statins, £ (prior to June 2012)247
Source: www.scribd.com/doc/122981341/Friday-Lunchtime-Lectures-at-The-ODI-Big-Data-Comes-to-the-NHS
The team’s decision highlights an important point: users of big data and public sector information
need to approach the analysis responsibly and in an informed way. The Mastodon C team
maintains that data itself is not a silver bullet to understanding a problem. The team needed to fully
understand the clinical guidance attached to each drug, so as to avoid, for example,
recommending savings where there were strong clinical reasons for choosing a branded drug
rather than the generic version. In their own words, “good domain knowledge usually beats supersmart algorithms.” In this case, while the data itself was crucial, the insights became possible only
through working closely with GPs and other healthcare professionals.
This issue of responsible use of the data also applied to the level of granularity of the analysis.
Although the data would allow comparisons to be made at a GP practice level, the Mastodon C
team opted to map their findings at a PCT/CCG248 level. This decision arose out of concerns that
presenting the results at a GP level could lead to distortions in the data involving very small
practices, as these might handle only a few relevant cases. This could mean that their prescribing
costs cost appear exceptionally high. The team were concerned that their work could lead to a GP
practice being unjustly labelled ‘the worst in Britain’ in the media. This might then create a backlash
and build resistance to future analyses and releases of data among the healthcare profession.
By comparing prescribing practice across England, the Mastodon C team identified around £200
million worth of potential savings. These could be achieved if the behaviour of the GPs with the
highest level of branded prescriptions was brought into line with the behaviour of the GPs
prescribing the highest level of generic versions. The mapping of results by PCT revealed notable
differences between PCTs across England.
247
Note: these prices have subsequently changed as Atorvastin came off patent in June 2012 and the price has now
dropped to £1.25 for 10mg: www.ppa.org.uk/ppa/edt_intro.htm. Prescribing Analytics reports that this was accounted for
in the analysis – see http://prescribinganalytics.com/analysis
248
Primary Care Trusts (PCTs) commission primary, community and secondary care from providers. Clinical
Commissioning Groups (CCGs) are the bodies that are to assume most of the commissioning responsibilities of PCTs
under the Health and Social Care Act 2012.
216
Market Assessment of Public Sector Information
Figure A7.2: percentage of proprietary statin prescribing by CCG Sept 2011 – May
2012
Source: http://prescribinganalytics.com/
The team also identified some potentially effective levers of influence over prescribing practice.
Cambridge PCT had very low prescribing costs for statins: the team found that this was because
the PCT had taken a very strong line on prescribing the generic version, and this had changed
behaviour at the GP practice level. The achievement of considerable cost savings in
Cambridgeshire demonstrates that action at the PCT level may be the most effective approach to
achieving savings through changes in GP prescribing behaviour.
It should be noted that work in this field is already undertaken by the government. For example, the
NHS Prescription Services, part of the NHS Business Services Authority, provides the NHS with a
range of drug, financial and prescribing information. The IT tools allow NHS organisations to look
at prescribing patterns for a range of medicines right down to individual GP level, so they can
target work with practices to help them improve. At a local level, NHS organisations have
developed systems to incentivise prescribers to prescribe more cost effectively. Many local NHS
organisations have also invested in software which prompts prescribers on the most cost effective
prescribing choice.249 These tools have helped the NHS to achieve one of the highest rates of
generic prescribing in Europe, with the overall prescribed generic rate calculated at 83 per cent in
2011.250
This highlights that in some instances where data is released, there may have been previous and
ongoing efforts by public sector organisations to use this data to develop insights. The work of
private sector organisations may be able to build on this work and enhance the value derived from
249
250
Based on Department of Health comments on an earlier draft
NHS Information Centre, Prescriptions Dispensed in the Community: England, Statistics for 2001 to 2011 (July 2012)
217
Market Assessment of Public Sector Information
the data, but this is likely to be most effective, and avoid duplication, where communications
between public and private sector organisations are strong.
In summary, the work of Mastodon C in this area demonstrates a number of key insights into the
value that may be derived from making more and better use of public sector information in the
health sector. These insights include:

it is not generally enough to have access to data – also required is the contextual and sector
knowledge to use it intelligently and responsibly;

large efficiency savings may be hidden in areas of regular practice. There are likely to be other
areas both in the NHS and the wider public sector where changes in behaviour could generate
significant financial savings without compromising outcomes;

the insights from data can help drive change in areas that previously resisted policy solutions.
For example, in this case a key insight was that change in GP prescribing behaviour might be
best driven from the PCT level;

there is a need for communication between public and private sector organisations to ensure
that where data is released, the analysis undertaken is complementary and builds on previous
and ongoing efforts to use the data to gain insights.
A patient database for the NHS – the challenges to extracting value from large datasets
Case study summary:
A central NHS patient database could offer significant savings to the NHS as well as improving standards of
care and the patient experience. However, there are significant technical, privacy and cultural hurdles to
overcome if this is to be made a reality. This case illustrates both how attractive the prize of harnessing the
power of large public sector information datasets can be, but also the difficulties these can present.
Size of the prize:
Difficult to estimate, but if the system is delivered as planned the savings are likely to be billions of pounds. An
initial illustrative estimate identified £4.4 billion of potential savings, although these are not all directly related
to the patient database.
The idea of a central NHS patient database has existed for some time, but previous attempts have
encountered a variety of issues owing to the challenges of building a system with the requisite
capabilities and scale. However, early in 2013 the Health Secretary, Jeremy Hunt, launched a plan
to store patient records in a cloud-based system by 2018, with the ultimate goal of transitioning
towards a ‘paperless NHS’.251
While the details have not been fully established, the broad outline of the scheme is that each
patient would have an individual electronic health record which would be accessible from any point
within the NHS – whether by the patient’s GP during a consultation, by a hospital consultant
preparing to carry out an operation, or an ambulance crew responding to an emergency call-out.
Outwardly at least these proposals have much to recommend them, and appear to be a good
example of an opportunity to achieve increased efficiency and improved outcomes through
increased exploitation of public sector information. Patient records offer a rich source of data with
the potential to both improve the standard of care received by the patient, and to enhance the
efficiency and effectiveness of the NHS. However, despite this promise, some have viewed the
proposals through the lens of past failures of NHS database projects, as well as privacy and
confidentiality concerns.
251
Jeremy Hunt’s announcement (16th January 2013) is available to view at:
http://mediacentre.dh.gov.uk/2013/01/16/16-january-2013-jeremy-hunt-policy-exchange-from-notepad-to-ipadtechnology-and-the-nhs/
218
Market Assessment of Public Sector Information
Commentators have also raised concerns that this database may threaten patient confidentiality. In
part this is simply a technical issue: a cloud-based, highly networked system will need to be
rendered secure against unwanted incursions by non-authorised agents, given that much of the
information held will be of a highly personal and sensitive nature.
If the information is released, either to approved users as is currently the case with the National
Pupil Database, there will be additional concerns regarding confidentiality252. As discussed below,
it may be possible to anonymise datasets to an extent that the level of risk of an individual record
becoming known is acceptable. However, healthcare information may pose particular challenges.
In the case of very rare diseases where, for example, there are only one or two cases per GP
practice area - it may be impossible to effectively anonymise the data. This indicates that the
approach to protecting patient privacy will need to be rigorous and adapted to the constraints of the
data.
In his announcement the Health Secretary acknowledged these concerns, but maintained that the
Government had learnt from past failures and would not be adopting a top-down approach in an
attempt to construct a monolithic central IT infrastructure. Instead of a centralised approach akin to
constructing an aircraft carrier, he argued that “most systems won’t necessarily need to be
replaced, just updated or adapted so they can talk with each other. A thousand different local
solutions linking together using common standards.”253
The announcement argued that this would deliver considerable financial benefits, based on a
report which identified potential savings amounting to £4.4 billion254. It should be noted that these
benefits were calculated from a wide range of changes which were estimated to deliver a range of
incremental benefits.
In addition, the Health Secretary’s announcement identified broader welfare benefits that include
improved outcomes, ultimately meaning lives saved, as well as an improved user experience of
engaging with the healthcare system at all levels. For example, if staff in Accident & Emergency
have details of a patient’s health history and current medication, they should be able to respond
more appropriately to the emergency, with potentially life-saving consequences. Equally, if
clinicians have access to up to date health records for all patients this means that patients would
not have to verbally repeat their medical history at each stage of the process. This is likely to save
time and ensure that clinicians have access to accurate records.
There is some international precedent for a centralised patient database. The most notable
example is Denmark, which currently makes hospital records available to patients online, and
which is in the process of making GP records available. However, Denmark has a population
numbering just ten per cent of the population of England alone and a population that is, prima
facie, more homogenous, meaning that a patient database in the UK will need to operate on a
much larger scale. This makes it a more challenging technical and logistical proposition.
This case throws into sharp relief the tensions around using large datasets concerning individuals.
On the one hand the potential for a range of benefits, from efficiency savings to enhanced user
convenience and new insights, is extremely tempting. On the other hand there are valid concerns
around privacy and confidentiality and the potential misuse of sensitive information. Effectively
harnessing the potential of data to transform the NHS, as well as other areas of the public sector,
will require these tensions to be reconciled in a manner which satisfies security and privacy
concerns while nonetheless permitting effective use of the data.
252
Needless to say, where release of personal information outside the NHS is intended, full consent of the patient(s)
concerned will be needed. Without such safeguards, it will be difficult to secure clinician and patient support for such a
database. This is particularly the case given that the information could affect the ability of patients to secure health
insurance, employment, and subject them to marketing attention, if released.
253
http://mediacentre.dh.gov.uk/2013/01/16/16-january-2013-jeremy-hunt-policy-exchange-from-notepad-to-ipadtechnology-and-the-nhs/
254
This can be accessed at www.wp.dh.gov.uk/publications/files/2013/01/Review-of-use-of-Information-andTechnology.pdf
219
Market Assessment of Public Sector Information
Publication of mortality rates following cardiac surgery
Case study summary:
Publishing data on mortality rates following adult cardiac surgery appears to be associated with a decline in
mortality. There are various theories as to why this is the case, including competitiveness among surgeons
which leads to a rise in performance, increased awareness among healthcare professionals, and public
pressure for higher standards. However, there are also concerns that the apparent decline in mortality may
reflect ‘gaming’ of the mortality data.
Size of the prize:
A decline in mortality rates is itself desirable. The economic value depends on the value attributed to a
statistical life. Taking a median value from a range of recent studies, the value of lives saved among those
undergoing adult coronary artery surgery in NHS centres in north-west England in 2005 was around £55
million. This suggests that the total value of lives saved per annum could exceed £400 million for England
and Wales, if similar benefits were observed in all regions.
Since 2005, data on mortality rates following adult cardiac surgery has been made publically
available in the UK. Subsequently, studies have attempted to gauge the impact that the release of
this data has had, both on outcomes of surgery and the willingness of surgeons to accept high-risk
cases.
There is a reasonable volume of evidence suggesting that publication of this data is associated
with a decline in mortality. A 2007 research paper published in the journal Heart found that over an
eight year period (from 1997 to 2005) observed mortality decreased from 2.4 per cent to 1.8 per
cent. The authors argue that while data was not made publically available over the majority of the
period covered by the study, over this period it became clear to surgeons that public release of the
data was a matter of time, following the Bristol Public Enquiry in 2001.255 The findings of the paper,
moreover, appear to be corroborated by other studies, including a study which found a 41 per cent
reduction in risk-adjusted mortality rates during the first four years following publication of
outcomes data.256
An improvement in clinical outcomes that leads to an increase in lives saved is of itself desirable
from a wider social welfare perspective. The economic value attributed to this decline in mortality
from surgery depends on the value of a statistical life, which varies widely. To take an example,
drawing on a survey of recent studies257 produces a median value for the total value of a statistical
lifetime of around £3 million. Using this value means that the decline in mortality can be valued at
around £55 million, just for north-west England in 2005. Assuming similar benefits were
experienced across England and Wales, the total value of lives saved in 2005 alone is estimated to
have exceeded £400 million, as compared with 1997 mortality rates. If similar benefits can be
realised in other areas of clinical practice, the annual benefits in terms of lives saved are likely to
be valued in terms of billions of pounds.
There have, however, been fears that public disclosure of outcomes could trigger risk-averse
behaviour among surgeons, meaning that they would not accept cases where there was an
increased risk of mortality.258 There is little evidence that this has in fact occurred, with the Heart
study finding that the number of patients classified as high-risk actually increased over the period.
However, a response to the article argued that surgeons have a powerful incentive to ‘game’ the
system by over-assessing the risk profile of patients, thereby making their own performance
255
See www.ncbi.nlm.nih.gov/pmc/articles/PMC1955202/#ref16
See Hannan E L, Kilburn H, Jr, Racz M. et al Improving the outcomes of coronary artery bypass surgery in New York
State. JAMA 1994
257
See The Government Office for Science, ‘Reducing Risks of Future Disasters: Priorities for Decision Makers’ (2012)
256
258
www.ncbi.nlm.nih.gov/pmc/articles/PMC1955009/
220
Market Assessment of Public Sector Information
appear stronger.259 There is some evidence that similar effects may have occurred elsewhere, with
one survey cited in the Heart study finding that 79 per cent of New York cardiologists reported that
publication of mortality statistics had influenced their decision about whether to perform angioplasty
on individual patients.260
These concerns notwithstanding, the evidence seems to indicate that transparency can be a
powerful driver of accountability and improved standards in the healthcare sector. As the authors of
the Heart paper suggest, “if public disclosure can drive data collection and analysis, but does not
create significant risk‐averse behaviour, its introduction may be beneficial in other areas of
medicine.”261 Releasing data on standards, where appropriate and subject to the appropriate
monitoring mechanisms, therefore appears to be an ‘easy win’ for the use of public sector
information to drive tangible economic and welfare benefits.
The public sector
One of the most interesting areas for the improved exploitation of public sector information is
greater sharing of and access to information within the public sector itself. By removing barriers to
the flow of information between public sector bodies – either because potential users are unaware
of what information is available, or because there are physical or legal constraints on the sharing of
this information – policy formation could be based on much richer and more complete information.
All things being equal, this should lead to better informed and therefore more effective policies.
A report by the Administrative Data Taskforce published in December 2012 highlighted a number
of areas of research and policy where sharing of data could generate new insights and lead to
improved policy outcomes. These are summarised below.
Area of research and policy
Description
Social mobility
Linking data on education, training, employment, unemployment,
income and benefits
Causal pathways over the life
course
Linking data on education, health, employment, income and wealth
Support for the elderly
Comparative analysis of access to and provision of social care support
for the elderly
Poverty
Linking data on housing conditions, health, incomes and benefits
Social care for children
Linking indicators of parental employment, social background and
childcare
259
260
Ibid.
See Hannan E L, Siu A L, Kumar D. et al Assessment of coronary artery bypass graft surgery performance in New
York. Is there a bias against taking high‐risk patients? Med Care 1997
261
See www.ncbi.nlm.nih.gov/pmc/articles/PMC1955202/#ref16
221
Market Assessment of Public Sector Information
Area of research and policy
Description
Offence and re-offence
Linking data on offending and re-offending behaviour, income,
benefits, health and mental health
Source: The UK Administrative Data Research Network: Improving Access for Research and Policy (December 2012)
This list serves as an indication of the many policy areas which could benefit from increased
sharing of information between government departments. Given the level of government spending,
and the wider economic significance of these policy areas, if sharing leads to enhancement of
policy in any one of these areas the economic and broader welfare impacts should be large. Below
is an example of one area in which progress is already being made in the sharing of information.
Announced in May 2012, the Social Mobility Transparency Board is tasked with pursuing “smarter
use of data between the Department for Education, the Department for Business, Innovation and
Skills (BIS) and HM Revenue and Customs.”262 The role of the Board is to improve sharing
between key information holders and external researchers, building connections and establishing
procedures to ease access to data for research.
The value of public sector information to social mobility researchers
From conversations with key stakeholders it is clear that public sector information is crucial for
researchers focussing on social mobility. Hitherto, most work in this field has been built upon birth
cohort studies which follow a group of subjects and update key indicators every ten years.
Data held by public sector bodies holds the promise of far richer and more robust insights into key
indicators of social mobility, potentially extending over the entire population. For researchers to use
this data effectively, however, it needs to be shared and linked across the various information
holders.
The organisations so far identified as holding the information most useful to researchers in this field
are:

the Department for Education;

the Higher Education Statistics Agency;

the Department of Work and Pensions; and

Her Majesty’s Revenue and Customs.
This being a relatively recent initiative, tangible benefits are still emerging. However, stakeholders
point to a number of incremental benefits. There is steady improvement in the level of
understanding of the drivers of social mobility, especially how higher education works as a driver of
social mobility and understanding of the factors driving participation in higher education. Insights
into the role of socio-economic group, school type, family income and other factors in influencing
participation in higher education continue to deepen as a result of increasing access to data.
Opportunities in the local public sector
The local public sector spends around £70 billion per year and employs approximately two million
people, providing many of the essential services across the UK. It therefore seems reasonable to
expect that there will be many opportunities to realise efficiencies and improved outcomes through
the more effective exploitation and sharing of information.
The local public sector also poses unique challenges owing to its great diversity. It includes county
and district councils (in two-tier ‘shires’), unitary authorities, London and metropolitan boroughs,
and sui generis authorities such as the City of London and Isles of Scilly. These administer a
262
See Open Data White Paper, June 2012
222
Market Assessment of Public Sector Information
diverse range of services under a variety of delivery models. Any changes in the use of public
sector information at a local level is therefore likely to be incremental and involve, at least initially,
either individual local authorities or small groups working together.
This diversity, however, is also a significant strength, since it makes local authorities laboratories
for the use and re-use of information. In the course of our research Deloitte has encountered many
examples of local authorities exploiting information in ways that are often highly innovative, with
lessons not only for other local authorities but also for central government departments and other
public sector bodies. The individual efficiency gains and improvements in outcomes may in many
cases be small, but collectively they represent a quiet revolution in the delivery of services at a
local level. Should these experiments grow in number and the best examples become widely
adopted, the national impact could be significant.
There are also numerous other examples of local authorities using information in innovative ways,
including:




Lambeth Borough Council have set up a site called “Lambeth in Numbers” to help inform
their Food Strategy work. It brings together data from various sources including central and
local government on a map.
Bristol City Council Air Quality data
Trafford Council – Breakthrough Fund on stimulating data sets
Hampshire County Council work on land supply
The example below illustrates an example of improvements in policy formation and delivery
through information sharing between local authorities and other public bodies.
An example of data sharing at a local level – families with complex needs
Examples of data sharing between local councils and other public bodies are proliferating across the
UK. These are often responses to the twin pressures of deep funding cuts and intractable problems
involving multiple agencies.
An example of this sort of problem is the case of families with complex needs, often referred to in the
press as ‘troubled families’. These families may combine issues such as mental health problems,
children out of school, and long term worklessness and benefit dependency, meaning that they fall
within the remit of multiple public sector bodies including social services, the Police, and the local
and national welfare services. These services are estimated to cost around £75,000 per family per
year.263
Increased coordination between local councils and other local partners may prove both more cost
efficient and more effective in resolving the problems faced by such families. The councils of Greater
Manchester, Leicestershire and Bradford are working together to improve information sharing and
management in this area.
The project aims to develop a single toolkit for information sharing, combining existing guidance and
approaches. This is intended to be applicable to both the councils and the agencies working with
families with complex needs.
In addition the project is intended to lay the foundations for a culture more conducive to information
sharing, addressing issues such as different professional cultures of sharing, lack of training and
expertise, and differing interpretations of legislation.
In practice, the steps needed to achieve this can appear prosaic but are nonetheless potentially
powerful enablers of an environment in which information can more easily be shared within and
between organisations.
As the project is still ongoing it is too early to judge its effects, whether in terms of reduced costs or
263
Association of Greater Manchester Authorities, ‘Improving Information Sharing and Management: A National
Exemplar Project’
223
Market Assessment of Public Sector Information
An example of data sharing at a local level – families with complex needs
improved outcomes. Nonetheless, this appears to be a positive example of the potential of greater
sharing and exploitation of data between councils and other public sector bodies in response to a
policy problem which has proved unresponsive to a siloed approach.
The Local Government Transparency Survey
The Local Government Association (LGA) conducted a survey in September 2012264 covering the
publication of open data, the impact of open data on councils and how data is used locally. 37 per
cent of local authorities responded to the survey. In their responses they identified the ways in
which they are using open data, which give an indication of the impact it is having. A selection of
these responses is included below.

engaging with community groups to create API makers and ways to make data useful;

using their own open data and that of other councils to gain insight into the characteristics of
people in the area and, accordingly, their needs in relation to the services the council currently
and intends to provide or commission;

planning to utilise the Police, NHS, public health and other public sector partners’ open data to
produce a single view of the borough;

seeing a reduction in FOI requests;

using data to help customers ‘self-serve’ online in their reporting e.g. fly tipping, tree issues,
damage to street furniture; and

undertaking service reviews and improvement – benchmarking, performance, spend,
organisational structures and pay scales.
These responses indicate the diversity of ways in which local authorities are making use of public
sector information from a wide range of sources, in order to achieve efficiency savings and improve
service delivery and the ways in which citizens interact with local government.
That said, the survey also identified barriers to local authority use of data. The barriers most
commonly cited by respondents were:

lack of resources to prepare and publish data (69 per cent);

issues around data protection and the release of personal information (32 per cent);

organisational and cultural barriers (30 per cent);

technical barriers (30 per cent); and

lack of skills to prepare and publish data (27 per cent).
Respondents also cited a range of other barriers, including:

a lack of a definitive list to help them know what information should be published;

a lack of clear guidance on data standards; and

a lack of mature offerings from suppliers to provide open data as part of standard operations.
Given how diverse the landscape is, in terms of both the extent to which local authorities are using
and releasing public sector information, the ways in which they are using it, and the barriers they
264
See Local Government Association, ‘Local Government Transparency Survey 2012’
224
Market Assessment of Public Sector Information
are facing, it makes sense to build on current efforts to share best practice across the local public
sector. This would provide a way to capitalise on local authorities as ‘laboratories’ for the use of
information. Where one local authority achieves efficiency savings or improves services, this may
be applicable to many other local authorities and in this way the benefits can be scaled up. It would
seem sensible to have a discussion about how and through what channels this can be achieved.
The LGA has already begun some work in this area, along with DCLG, through initiatives such as
the Local E-Government Standards Board (LEGSB) and the Local Public Data Panel. There
appears to be support for such initiatives from local authorities themselves: in the survey, around
two thirds were in favour of the LGA providing a framework for publishing data through by way of a
transparency strategy, and indicated that they would like to see case studies and guidance on how
to publish open data.
Summary of opportunities in the local public sector
The examples given above indicate some of the ways in which the local public sector is beginning
to exploit information to deliver both efficiencies and improved outcomes for policies and services,
leading to welfare gains for citizens. These examples are likely to be a leading edge for
significantly expanded future benefits, if the current trend of increasing sharing and exploitation of
data continues.
The report by ConsultingWhere and ACIL Tasman on the use of geospatial information in the local
public sector identified the following benefits for 2009:

GDP was £323m higher than it would otherwise have been (an increase of 0.02 per cent);

government revenue from taxation was £44m higher than it would otherwise have been;

the delivery of goods and services by local public service providers was £232m higher than it
would otherwise have been; and

an increase in labour productivity equivalent to 1,500 full time staff across England and Wales,
owing to the effects of improved citizen and business contact with local service providers.265
These benefits have arisen from the use of just one type of information (geospatial) by some, but
not all local authorities. Given that the local public sector collectively accounts for expenditure of
some £70 billion per year266, and is often the primary point of contact for citizens and businesses
dealing with issues as diverse as education, social services and planning permission, the potential
for further gains – both to local public sector efficiency and the wider economy – is likely to be
significant.
As this section has demonstrated, the strength of the local public sector is its diversity, making it a
laboratory for the use of information. In order to facilitate the potential of information release and
sharing, it is important to ensure that the necessary frameworks (legal, regulatory and in terms of
culture) are in place, and that local authorities have the freedom to experiment while also receiving
suitable guidance. Under these conditions the potential for public sector information to transform
local public service delivery and efficiency is likely to be significant.
265
ACIL Tasman and ConsultingWhere, ‘The Value of Geospatial Information to Local Public Service Delivery in England
and Wales’ (July 2010)
266
Ibid.
225
Appendix 8: Government Officials
informal consultation
Consultation questions
Background
Deloitte has been engaged by the Data Strategy Board to conduct a market assessment of Public
Sector Information (PSI). The results of study will feed into the independent Shakespeare Review
which is examining ways to widen access to Public Sector Information (PSI) and consider new and
innovative opportunities for it.
Survey
1. Please could you provide your name; organisation; job title; role in the collection/dissemination
of PSI in your organisation; and your contact details (telephone and email address)?
2. Approximately what proportion of the data that your organisation collects is made available to
the general public (either freely or for a fee)?. Choose one option only.
a. None
b. Between 0 and 25%
c. Between 26 and 50%
d. Between 51 and 75%
e. More than 75%
f.
All data is made available
g. Don’t know
3. Approximately what proportion of the data made available to the general public by your
organisation is at no cost? Choose one option only.
a. 100% is free
b.
Between 75 and 99% is free
c. Between 50 and 74% is free
d. Between 25 and 49% is free
e. Less than 25% is free
f.
None is free
g. Don’t know
4. How does your organisation make data available to the general public? Indicate all that apply.
a. Own website
b. On data.gov.uk
c. On another data portal (please specify)
d. Upon receipt of special requests
e. Other (please specify)
5. Approximately what proportion of your organisation’s staff are directly involved in the collection,
processing and dissemination of data? Choose one option only.
Market Assessment of Public Sector Information
a. Less than 5%
b. Between 5 and 10%
c. Between 11 and 25%
d. More than 25%
e. Don’t know
6. Approximately what proportion of your organisation’s budget is spent on the collection,
processing and dissemination of data? Choose one option only.
a. Less than 5%
b. Between 5 and 10%
c. Between 11 and 25%
d. More than 25%
e. Don’t know
7. Would you say that the data collected by the department is primarily used by: Choose one
option only.
a. Your own organisation
b. Other Government organisations and agencies
c. Third parties – please specify
d. Don’t know who the main users / re-users of data are
8. What sort of requests do you receive from the general public regarding open data? Indicate all
that apply.
a. Requests for other data that your organisation collects to be made public
b. Requests for your organisation to provide new data that is currently not collected
c. Requests for clarification of various issues around datasets or to improve the quality
of datasets
d. Requests to correct different aspects of datasets
e. Requests for the data currently available publicly to be made available in different
formats
f.
Other – please specify
9. Are you able to give an indication of the types of users downloading your organisation’s
datasets? Indicate all that apply and provide an approximation of the proportion of users they
represent.
a. Other Government organisations and agencies
b. Other public service providers
c. Not-for profit organisations including researchers
d. For-profit private sector organisations
e. Individuals
f.
Other – please specify
10.
Can you given indication of how the data is used and re-used?
11.
Does your organisation plan to make available more data to the general public in the near
future (within twelve months)?
227
Market Assessment of Public Sector Information
a. Yes
b. No
c. Don’t know
12.
If the above answer is yes, how much more data does your organisation plan to make
available to the general public?
a. Less than 25%
b. Between 26 and 50%
c. Between 51 and 75%
d. More than 75%
e. Don’t know
13.
What are the key challenges you see in making more data available to the general public?
Consultation responses
The following organisations responded to the survey:












HMRC
Home Office
Defra
DH
DH Statistics Function
DH MHRA
FCO
DCLG
MoD
MoJ
CO
BIS
The following table shows the responses to each questions, by percentage of respondents
selecting each option. Note that in some cases it was possible to select more than one response,
meaning that the sum of responses for these questions exceeds the total number of respondents.
Question
a
b
c
d
e
f
g
1
n/a
n/a
n/a
n/a
n/a
n/a
n/a
2
8%
25%
0%
0%
17%
8%
33%
3
42%
42%
0%
0%
0%
0%
8%
4
92%
83%
67%
75%
33%
0%
0%
5
33%
0%
0%
8%
50%
0%
0%
6
42%
0%
0%
0%
50%
0%
0%
7
50%
8%
17%
33%
0%
0%
0%
8
67%
50%
75%
33%
58%
8%
0%
9
50%
42%
50%
42%
42%
33%
0%
10
n/a
n/a
n/a
n/a
n/a
n/a
n/a
11
83%
8%
0%
0%
0%
0%
0%
12
42%
8%
0%
8%
25%
0%
0%
228
Market Assessment of Public Sector Information
Question
a
b
c
d
e
f
g
13
n/a
n/a
n/a
n/a
n/a
n/a
n/a
229
Appendix 9: Bibliography
Academic Papers & Reports
ACIL Tasman, 2009. Spatial Information in the New Zealand Economy. Wellington: Land Information
New Zealand.
Administrative Data Taskforce, 2012. The UK Administrative Data Network: Improving Access for
Research and Policy. London: Economic and Social Research Council.
Association of Greater Manchester Authorities. Helping to improve information sharing and management:
a national exemplar project.
Capgemini Consulting, 2013. The Open Data Economy: Unlocking Economic Value by Opening
Government and Public Data. London: Capgemini Consulting.
Cebr, 2012. Data equity: unlocking the value of big data. London: SAS.
Cebr, 2012. The Economic Costs of Gridlock. London: Cebr.
Centre for Technology Policy Research, 2010. Open Government: some next steps for the UK.
ConsultingWhere and ACIL Tasman, 2010. The Value of Geospatial Information to Local Public Sector
Service Delivery in England and Wales. London: Local Government Association.
Danish Enterprise and Construction Authority, 2010. The Value of Danish Address Data. Copenhagen.
Dekkers, M., Polman, F., te Velde, R. and de Vries, M., 2006. Measuring European Public Sector
Information Resources. Helm Group of Companies.
Deloitte Access Economics, 2012. The Economic Impact of the “Information Glut”. London.
Deloitte Analytics, 2012. Open data: driving growth, ingenuity and innovation. London.
Deloitte Consulting and Tech4i2, 2011. Pricing of Public Sector Information Study.
Department for Transport, 2011. The Economic Case for HS2: The Y Network and London –West
Midlands. London: Department for Transport.
Department for Transport, 2012. Road Transport Forecasts 2011. London: Department for Transport.
Department for Transport, 2012. Values of Time and Vehicle Operating Costs. London: Department for
Transport.
Earnst & Young, 2011. Making Energy Efficiency Your Business: understanding the potential of the nondomestic Green Deal. London: Earnst & Young.
Elgin, 2013. A new public-private model for creating a national database of local roadworks
Market Assessment of Public Sector Information
e-skills UK, 2013. Big Data Analytics: an assessment of demand for labour and skills, 2012-2017.
London: SAS UK.
Genovese, E., Roche, S., Caron, C. and Feick, R., 2010. The EcoGeo Cookbook for the assessment of
Geographic Information Value. International Journal of Spatial data Infrastructure Research.
Goodwin, P., 2004. The Economic Costs of Road Traffic Congestion. London: University College London.
Graves, A. The Price of Everything but the Value of Nothing. Office of Fair Trading presentation.
Heseltine, The Rt Hon the Lord of Thenford CT, 2012. No Stone Unturned in Pursuit of Growth.
HM Government, 2012. Open Data White Paper: Unleashing the Potential. London.
Houghton, J., 2011. Costs and Benefits of Data Provision. Victoria: Centre for Strategic Economic
Studies, Victoria University.
Huijboom, N. and van der Broek, T., 2011. Open data: an international comparison of strategies.
European Journal of ePractice.
Kalapesi, C., Willersdorf, S. and Zwillenberg, P., 2012. The Connected Kingdom: how the internet is
transforming the UK economy. London: Boston Consulting Group, Inc.
Kundra, V., 2012. Digital Fuel of the 21st Century: Innovation Through Open Data and the Network Effect.
Joan Shorenstein Center on the Press, Politics and Public Society.
Lippert, C., 2010. Public Sector Information Reuse in Denmark. European Public Sector Information
Platform Report No. 20.
Local Government Association, 2012. Local Government Transparency Survey 2012. London: Local
Government Association.
Mandela, A., 2009. PSI re-use: UK perspectives. Locus Association.
Mayo, E. and Steinberg, T., 2007. The Power of Information: An Independent Review.
McKinsey Global Institute, 2011. Big data: the next frontier for innovation, competition, and productivity.
McKinsey & Company.
National Audit Office, 2012. Implementing transparency: cross-government review. London: National
Audit Office.
National Fraud Authority, 2012. Annual Fraud Indicator. London: National Fraud Authority.
O’Hara, K., Whitley, E. and Whittal, P., 2011. Avoiding the jigsaw effect: experiences with Ministry of
Justice reoffending data. University of Southampton, London School of Economics, and Detica.
Office of Fair Trading, 2006. The Commercial Use of Public Information. London: Office of Fair Trading.
Open Data User Group, 2013. Further Benefits of an Open National Address Set. London: Open Data
User Group.
Oxera, 1999. The Economic Contribution of Ordnance Survey GB. Oxford: Oxera.
231
Market Assessment of Public Sector Information
PA Consulting Group, 2007. Met Office: The Public Weather Service’s Contribution to the UK Economy.
London: PA Consulting Group.
Pira International, 2000. Commercial Exploitation of Europe’s Public Sector Information
Pollock, R., 2011. Welfare Gains from Opening up Public Sector Information in the UK. Cambridge:
University of Cambridge
Pollock, R., Newbury, D. and Bently, L., 2008. Models of Public Sector Information Provision via Trading
Funds. Cambridge: University of Cambridge.
PriceWaterhouseCoopers LLP, 2010. Economic Assessment of Spatial Data Pricing and Access.
Canberra: ANZLIC – The Spatial Information Council.
PriceWaterhouseCoopers LLP, 2013. A review of the potential benefits from the better use of information
and technology in Health and Social Care. London: PriceWaterhouseCoopers LLP.
Public Data Group. Business Case.
Roger Tym & Partners, 2003. The Economic Benefit of the British Geological Survey. London: Roger
Tym & Partners.
Rothenberg, J., 2012. Towards a better supply and distribution process for open data: international
benchmark. Logica Business Consulting.
Stauffacher, D., Hattotuwa, S. and Weekes, B., 2012. The potential and challenges of open data for crisis
information management and aid efficiency: a preliminary assessment. ICT4Peace Foundation.
The National Archives, 2009. How we deal with exceptions to marginal cost pricing. London: The National
Archives.
The National Archives, 2010. The United Kingdom Report on the Re-use of Public Sector Information
2010: unlocking PSI potential. London: The National Archives.
The Office of Public Sector Information, 2011. Information Fair Trader Scheme Report: Ordnance Sur
vey. London: The National Archives.
The Office of Public Sector Information, 2012. Information Fair Trader Scheme Report: Land Registry.
London: The National Archives.
Transport for London, 2011. Travel in London, Supplementary Report: London Travel Demand Survey.
London: Transport for London.
Transport for London, 2011. Travel in London: Report 5. London: Transport for London.
Transport for London, 2012. London Buses Performance: Financial Year 2011/12. London: Transport for
London.
Transport for London, 2012. London Underground Performance Report: Period 4 2012/13. London:
Transport for London.
Transport for London, 2012. Syndication Developer Guidelines: Transport for London Data Service
Version 2.0. London: Transport for London.
232
Market Assessment of Public Sector Information
Transport for London: London Streets, 2012. Performance Report: Quarter 1 2012/13. London: Transport
for London.
Uhlir, P. F., 2009. The Socioeconomic effects of public sector information on digital networks: towards a
better understanding of different access and reuse policies (workshop summary). US National Committee
for CODATA, and Board on Research Data and Information.
Vickery, G., 2011. Review of recent studies on PSI re-use and related market developments. Information
Economics.
Yiu, C., 2012a. A Right to Data: Fulfilling the Promise of Open Public Data in the UK. London: Policy
Exchange.
Yiu, C., 2012b. The Big Data Opportunity: Making government fast, smarter and more personal. London:
Policy Exchange.
Departmental Open Data Strategies
Cabinet Office, 2012. Open Data Strategy. London: Cabinet Office.
Defra, 2012. Open Data Strategy June 2012 – March 2014. London: Defra.
Department for Business, Innovation & Skills, 2012. Open Data Strategy 2012-14. London: Department
for Business, Innovation & Skills.
Department for Communities & Local Government, 2012. Open Data Strategy April 2012-April 2014.
London: Department for Communities & Local Government.
Department for Culture, Media and Sport, 2012. Open Data Strategy 2012-15. London: Department for
Culture, Media and Sport.
Department for Education, 2012. Open Data Strategy. London: Department for Education.
Department for International Development, 2012. Open Data Strategy April 2012 – March 2014. London:
Department for International Development.
Department for Transport, 2012. Open Data Strategy. London: Department for Transport.
Department for Works and Pensions, 2012. Open Data Strategy. London: Department for Works and
Pensions.
Department of Energy and Climate Change, 2012. DECC’s Open Data Strategy. London: Department of
Energy and Climate Change.
Department of Health, 2012. The Power of Information: putting all of us in control of the health and care
information we need. London: Department of Health.
Foreign and Commonwealth Office, 2012. The FCO’s Open Data Strategy. London: Foreign and
Commonwealth Office.
HM Revenue & Customs, 2012. Open Data Strategy. London: HM Revenue & Customs.
HM Treasury, 2012. Open Data Strategy 2012-2015. London: HM Treasury.
233
Market Assessment of Public Sector Information
Home Office, 2012. Open Data Strategy for the Home Office 2012-13. London: Home Office.
Ministry of Defence, 2012. Open Data Strategy 2012-2014. London: Ministry of Defence.
Ministry of Justice, 2012. Open Data Strategy 2012-15. London: Ministry of Justice.
Other Documents
Association of Train Operating Companies, 2012. RSP licence fee and datafeeds charge.
Cabinet Office, 2012. Making open data real: a government summary of responses. London: Cabinet
Office.
Department of Business Innovation and Skills, 2012. Annual Report and Accounts 2011-2012. London:
Department of Business Innovation and Skills.
HM Government, 2011. Making open data real: a public consultation. London: HM Government.
Met Office, 2012. Annual Report and Accounts 2011/12. Exeter: Met Office.
Open Data User Group, 2012. Open Data User Group response to the consultation on the code of
practice for datasets and beta charged licence. London: Open Data User Group.
234
Market Assessment of Public Sector Information
Crown copyright 2013
You may re-use this information (not including logos) free of charge in any format or medium, under the terms of the
Open Government Licence. Visit www.nationalarchives.gov.uk/doc/open-government-licence, write to the
Information Policy Team, The National Archives, Kew, London TW9 4DU, or email:
psi@nationalarchives.gsi.gov.uk.
This publication is available from www.gov.uk/bis
Any enquiries regarding this publication should be sent to:
Department for Business, Innovation and Skills
1 Victoria Street
London SW1H 0ET
Tel: 020 7215 5000
If you require this publication in an alternative format, email enquiries@bis.gsi.gov.uk, or call 020 7215 5000.
BIS/13/743
235